Reactions of Saturated and Unsaturated Tertiary Alkyl Halides and Saturated Secondary Alkyl Iodides with Lithium Aluminum Deuteride. Convincing Evidence for a Single-Electron-Transfer Pathway

Article abstract of DOI:10.1021/jo9807058

Reactions of saturated secondary and tertiary alkyl halides with LiAlH4 (LAH) and LiAlD4 (LAD) have been carried out, and convincing evidence for a single-electron-transfer (SET) pathway has been obtained. Reactions involving saturated alkyl halides with LAD provide a model system in which a halogen-atom radical chain process is not possible, and therefore, the observation of large quantities (in some cases >90 percent) of protium in the reduction product provides strong evidence for a radical intermediate and a SET pathway. Specifically, the reaction of 2-iodooctane (10) with LAD produced octane with a deuterium content as low as 21 percent. Also, the reaction of the tertiary halide 2-iodo-2-methylheptane (15) with LAD produced 2-methylheptane (16) with a deuterium content as low as 8 percent. The effect of stoichiometry, halogen type, and reaction vessel surface on these reactions was studied. Reaction of the unsaturated tertiary halide 6-iodo-6-methyl-1-heptene (25) with LAD was also studied and was found to proceed predominantly by a SET process involving a halogen-atom radical chain process. The possibility of a carbocation intermediate in all of these reactions is discussed.

Full text of DOI:10.1021/jo9807058

J . Org. Chem. 1998, 63, 7707-7714
7707
Rea ction s of Sa tu r a ted a n d Un sa tu r a ted Ter tia r y Alk yl Ha lid es
a n d Sa tu r a ted Secon d a r y Alk yl Iod id es w ith Lith iu m Alu m in u m
Deu ter id e. Con vin cin g Evid en ce for a Sin gle-Electr on -Tr a n sfer
P a th w a y
E. C. Ashby* and Catherine O. Welder
School of Chemistry and Biochemistry, Georgia Institute of Technology, Atlanta, Georgia 30332-0400
Received April 15, 1998
Reactions of saturated secondary and tertiary alkyl halides with LiAlH4 (LAH) and LiAlD4 (LAD)
have been carried out, and convincing evidence for a single-electron-transfer (SET) pathway has
been obtained. Reactions involving saturated alkyl halides with LAD provide a model system in
which a halogen-atom radical chain process is not possible, and therefore, the observation of large
quantities (in some cases >90%) of protium in the reduction product provides strong evidence for
a radical intermediate and a SET pathway. Specifically, the reaction of 2-iodooctane (10) with
LAD produced octane with a deuterium content as low as 21%. Also, the reaction of the tertiary
halide 2-iodo-2-methylheptane (15) with LAD produced 2-methylheptane (16) with a deuterium
content as low as 8%. The effect of stoichiometry, halogen type, and reaction vessel surface on
these reactions was studied. Reaction of the unsaturated tertiary halide 6-iodo-6-methyl-1-heptene
(25) with LAD was also studied and was found to proceed predominantly by a SET process involving
a halogen-atom radical chain process. The possibility of a carbocation intermediate in all of these
reactions is discussed.
In tr od u ction
chose saturated probes in order to eliminate the possibil-
ity of a halogen-atom radical chain process that is
available to the unsaturated, cyclizable probes (Scheme
1). It has been argued that the halogen atom radical
As early as 1984, we reported evidence for extensive
cyclization in the reaction of the hindered alkyl iodide
radical probe 6-iodo-5,5-dimethyl-1-hexene (1) with Li-
AlH4 (LAH; eq 1) and proposed that this reaction pro-
Sch em e 1
ceeds by a single-electron-transfer (SET) process involv-
ing a radical intermediate.1 More recently we have
reported that the saturated alkyl iodide probes, namely,
the hindered primary probe 1-iodo-2,2-dimethylhexane
(4)2 and the unhindered primary probe 1-iodooctane (6),3
react with LiAlD4 (LAD) to form substantial amounts of
the nondeuterated reduction products 2,2-dimethylhex-
ane (5) and n-octane (7), respectively (eqs 2 and 3). We
chain process shown in Scheme 1 can be initiated by
unknown impurities,4 and therefore, the amount of
cyclized product, indicative of a radical precursor, cannot
be used to quantitate the amount of initial electron
transfer from the nucleophile.4,5 Removal of the double
bond in the alkyl halide probes results in removal of the
possibility of a halogen-atom radical chain process,
making halogen-atom exchange between radical inter-
mediates and starting materials a nonproductive process.
As explained previously,2 the product of a SET reaction
of 4 with LAH would be indistinguishable from the SN2
product; therefore, LAD was necessarily chosen as the
nucleophile in the study of saturated probes so that in
(2) Ashby, E. C.; Welder, C. O. J . Org. Chem. 1997, 62, 3542.
(3) Welder, C. O.; Ashby, E. C. J . Org. Chem. 1997, 62, 4829.
(4) (a) Newcomb, M.; Curran, D. P. Acc. Chem. Res. 1988, 21, 206.
(b) Newcomb, M.; Kaplan, J . Tetrahedron Lett. 1988, 29, 3449. (c)
Newcomb, M.; Kaplan, J .; Curran, D. P. Tetrahedron Lett. 1988, 29,
3451. (d) Newcomb, M.; Sanchez, R. M.; Kaplan, J . J . Am. Chem. Soc.
1987, 109, 1195. (e) Park, S. U.; Chung, S. K.; Newcomb, M. J . Org.
Chem. 1987, 52, 3275.
(1) (a) Ashby, E. C.; Pham, T. N.; Amrollah-Madjdabadi, A. J . Org.
Chem. 1991, 56, 1596. (b) Ashby, E. C. Acc. Chem. Res. 1988, 21, 414.
(c) Ashby, E. C.; DePriest, R. N.; Goel, A. B.; Wenderoth, B.; Pham, T.
N. J . Org. Chem. 1984, 49, 3545. (d) Tolbert, L. M.; Sun, X. J .; Ashby,
E. C. J . Am. Chem. Soc. 1995, 117, 2681.
(5) Curran, D. P.; Kim, D. Tetrahedron Lett. 1986, 27, 5821.
10.1021/jo9807058 CCC: $15.00 © 1998 American Chemical Society
Published on Web 10/10/1998

7708 J . Org. Chem., Vol. 63, No. 22, 1998
Ashby and Welder
Ta ble 1. Effect of Stoich iom etr y on th e Rea ction of
2-Iod oocta n e (10) w ith LAD in THF a
the SN2 reaction of LAD with an alkyl halide only
deuterated alkane is expected. However, for a reaction
involving radical intermediates, hydrogen atoms (from
solvent) as well as deuterium atoms (from LAD and its
byproducts) may be abstracted so that both the nondeu-
terated and the deuterated alkane could be formed (eq
4).
material
exp RI/Nuc time, h % 10 % 11 (% D) % dimers balance
1
1:5
1
2
5
0.5
1
2
4
6
1
2
5
8
24
48
96
20
trace
0
62
45
25
9
79
89
trace
trace
trace
99
89
94
99
98
99
100
97
95
94
94
95
94
96
97
94 (95)
37 (87)
53 (86)
74 (84)
91 (82)
97 (81)
13 (52)
18 (46)
23 (33)
26 (29)
27 (27)
31 (23)
35 (21)
2
1:1
trace
trace
trace
trace
1
2
2
2
2
2
2
0
3
1:0.2
81
74
69
67
65
63
60
To further extend the scope of the reaction of LAH with
alkyl halides, we have studied the reaction of typical
tertiary alkyl halides with LAD. The only reactions of
tertiary alkyl halides with LAD that we reported previ-
ously involve the 1-halonorbornanes.6 In the 1-halonor-
bornane/LAD system, we found significant evidence for
SET, as evidenced by low deuterium incorporation in the
product (norbornane). We present herein the results of
the reactions of the saturated, noncyclizable tertiary
halides (2-halo-2-methylheptanes), the unsaturated, cy-
clizable tertiary halides (6-halo-6-methyl-1-heptenes),
and the saturated secondary alkyl halide (2-iodooctane,
10) with LAD.
a
All reactions were carried out at a RI concentration of 0.070
M in an Ar atmosphere glovebox in the absence of light at room
temperature in used Pyrex flasks and with the use of Teflon-coated
stir bars.
Ta ble 2. Effect of Rea ctor Su r fa ce on th e Rea ction of
2-Iod oocta n e (10) w ith LAD in THF a
reaction
vessel
time,
h
%
% 11
%
material
exp
1
RI/LAD
1:1
10 (% D) dimers balance
used Pyrex
0.5 62 37 (87) trace
99
98
1
2
4
6
45 53 (86) trace
25 74 (84) trace
99
9
0
91 (82) trace
97 (81) trace
100
97
2
quartz
1:1
0.5 82 18 (92)
100
97
93
82
81
Resu lts a n d Discu ssion
1
2
4
68 29 (89) trace
45 47 (84)
16 63 (78)
1
3
5
6
I. Rea ction s of Sa tu r a ted Secon d a r y Alk yl Ha -
lid es w ith LAD. Effect of Stoich iom etr y on th e
Rea ction of 2-Iod oocta n e (10) w ith LAD. The trends
of the reactions of secondary alkyl halides with LAD were
expected to follow the same patterns as those observed
for the primary systems discussed earlier.2,3 Therefore,
it was predicted that a decrease in the number of
equivalents of LAD present in the reduction of 10 should
lead to a decrease in the deuterium content of the
resulting alkane, if a radical mechanism was involved.
As reported in Table 1, when a 5-fold excess of LAD
was allowed to react with 10, no clear evidence of electron
transfer was obtained (experiment 1). A trace amount
of dimer formation was noted, and the 95% D found in
the octane was only a little low for the results expected
for a pure SN2 reaction. However, when the stoichiom-
etry of the reaction was reduced from a RI/LAD ratio of
1:5 to 1:1, after 6 h a significant amount of nondeuterated
alkane 11 (19% of the octane) was obtained (experiment
2). At a stoichiometry of 1:0.2, after 96 h, 79% of the
octane formed was nondeuterated. This trend is consis-
tent with radical formation via SET for the reaction of
10 with LAD as well as the stoichiometry studies
presented earlier for 1-iodooctane (6)3 and 1-iodo-2,2-
dimethylhexane (4).2
6
24
5
0
71 (70)
75 (67)
81
a
All reactions were carried out at a RI concentration of 0.070
M in an Ar atmosphere glovebox in the absence of light at room
temperature and with the use of Teflon-coated stir bars.
shown to have an effect on the reduction of the secondary
iodide 10 by LAD (Table 2). The reaction in a used Pyrex
vessel (experiment 1) proceeded somewhat more rapidly
than the reaction in quartz (experiment 2) carried out
under otherwise identical conditions. Also, it is evident
that the deuterium content in the octane was lower and
the dimer formation higher in quartz than in the used
Pyrex vessel, indicating a higher free radical content in
the quartz vessel. Because more radical formation was
observed in the reactions carried out in quartz, the
reactions carried out in Pyrex should represent a mini-
mum amount of products, indicative of a radical pathway.
Rea ction of 2-Br om oocta n e w ith LAD. Less SET
was expected for the reaction of 2-bromooctane (12) with
LAD than the corresponding iodide 10 because of the less
favorable reduction potential of alkyl bromides compared
to alkyl iodides. When alkyl bromide 12 was allowed to
react with LAD at 1:5, 1:1, and 1:0.2 ratios, no evidence
for SET was obtained.
P r op osed Mech a n ism of th e Red u ction of 2-Io-
d oocta n e (10) by LAD. As shown in Scheme 2, it is
suggested that alkyl iodide 10 can react with LAD by
either an SN2 or SET pathway. As shown in step a, LAD
can react in an SN2 fashion to yield 11d 1, AlD3, and LiI.
On the other hand, LAD can transfer an electron to 10
to form the corresponding radical anion 13 (step b) which
rapidly dissociates to form radical 14 (step c). Radical
Effect of Rea ctor Su r fa ce on th e Rea ction of
2-Iod oocta n e (10) w ith LAD. Because the nature of
the vessel surface was found to have an effect on the
reaction of the primary alkyl iodides with LAD,7 vessels
other than used Pyrex were employed in a study of the
reaction of 10 with LAD. The type of reactor surface was
(6) Ashby, E. C.; Sun, X.; Duff, J . L. J . Org. Chem. 1994, 59, 1270.
(7) For a study of the effect of reactor surfaces on SET reactions,
see: Ashby, E. C.; Welder, C. O. Tetrahedron Lett. 1995, 36, 7171.

Evidence for a SET Pathway
Sch em e 2
J . Org. Chem., Vol. 63, No. 22, 1998 7709
Ta ble 3. Effect of Stoich iom etr y on th e Rea ction of
2-Iod o-2-m eth ylh ep ta n e (15) w ith LADa
time,
h
%
15
% 16
%
%
%
%
material
exp RI/LAD
(% D) 17 18 19 dimers balance
1
1:5
0.5 47
2
5
5
7
16 30
20 13
35 15
100
98
97
95
99
95
87
87
91
94
1
2
5
57
33
8
3 (62)
7 (59)
17 (51) 10 51 9
24
1
5
8
24
48
trace 26 (47) 12 61 trace trace
2
1:0.2
81
60
49
40
35
2 (14)
11 (10)
15 (8)
15 (8)
16 (8)
2
6
7
8
9
4
5
8
5
6
5
3
trace
trace
2
3
4
8
20
27
a
All reactions were carried out at a RI concentration of 0.070
M in an Ar atmosphere glovebox in the absence of light at room
temperature in used Pyrex flasks and with the use of Teflon-coated
stir bars.
Table 3 reports the results of the reaction of 15 with
LAD carried out at two different stoichiometries (1:5 and
1:0.2). In both cases, significant evidence of SET was
observed, as noted by the low deuterium content in
alkane 16. Conversion of the absolute yields to relative
yields, based on the amount of starting material con-
sumed, shows that the percent of octane formed (26%)
was identical for experiments 1 and 2. However, only
8% D was found in the octane when 0.2 equiv of LAD
was present in the reaction as compared to 47% D when
equiv of LAD was present in the reaction. Therefore,
much more hydrogen-atom abstraction from the THF
took place in experiment 2 than in experiment 1, as
expected.
14 can then react with a suitable hydrogen-atom donor,
such as the solvent, to form octane (11) (step d), it can
combine with another alkyl radical 14 in solution to form
a dimer (step e), or it can abstract a deuterium atom from
-
AlD4 to form 2-deuteriooctane (11d 1) (step f) and the
AlD3 radical anion. The AlD3 radical anion can then act
as the major single-electron donor to the starting iodide
10 (step g) yielding radical anion 13 and AlD3, thus
producing a hydrogen atom radical chain process. The
AlD3 byproduct can then react with 10 (steps h-k) to
form 13 as shown.
There is a possibility of carbocation formation which
can lead to 16, 16d 1, 17, and 18 (eq 6). If carbocation
II. Rea ction s of Sa tu r a ted Ter tia r y Alk yl Ha lid es
w ith LAD. Tertiary alkyl halides are not expected to
undergo SN2 reactions; therefore, the SN2 pathway should
not compete with the SET pathway for tertiary systems.
However, other pathways (SN1, E1, and E2) can compete
with SET in the tertiary system. To study the competi-
tion between these pathways, several tertiary halides
were allowed to react with LAD.
Effect of Stoich iom etr y on th e Rea ction of 2-Iod o-
2-m eth ylh ep ta n e (15) w ith LAD. Five products were
identified in the reaction of tertiary iodide 15 with LAD
(eq 5). The observation that the percentage of alcohol
formation were the precursor to the formation of 16d 1
(step b, eq 6), then one would expect the amount of 16d 1
to increase (as compared to 16) as the ratio of LAD to 15
increases. J ust the opposite was observed. Another
pathway to 16 from a carbocation involves hydride ion
abstraction from THF (step c) which is not nearly as
energetically favorable as hydrogen atom abstraction
from THF by a radical intermediate. In a previous
study,6 we were able to show that the tertiary iodide
1-iodonorbornane, which is not easily capable of forming
a bridgehead carbocation,8 still reacts with LAD to form
predominantly the protio product (eq 7). Such a result
19 decreases as the starting alkyl iodide 15 is consumed
supports the hypothesis that alcohol 19 is formed in the
aqueous workup procedure (Table 3). The observation
that the percentage of alkenes 17 and 18 increases as
the percentage of 15 and 19 decreases implies that the
alkenes are produced during the reaction of 15 with LAD
and not solely in the aqueous workup. Therefore, only
four true products were obtained in the reaction of 15
with LAD: 16, 16d 1, 17, and 18.

7710 J . Org. Chem., Vol. 63, No. 22, 1998
Ashby and Welder
can only be explained by a reaction involving a free
radical intermediate.
proposed that LAD can react with tertiary halides in an
E2 fashion to give alkenes 17 and 18 (step a). LAD can
also act as a single-electron donor toward the alkyl
halides 15, 20, or 21 to form the corresponding radical
anion 22 (step b) from which products 16-18 are formed
(steps d-f).
Effect of Rea ctor Su r fa ce on th e Rea ction of
2-Iod o-2-m eth ylh ep ta n e (15) w ith LAD. In the previ-
ously reported reactions,7 quartz was found to be the
reaction vessel of choice for the unequivocal observation
of SET products. The reaction of 15 with LAD was
carried out in used Pyrex and compared to the same
experiment carried out in quartz (Table 4, expts 1 and
2). The reaction proceeds faster in quartz than in used
Pyrex. (Recall that alcohol 19 is a product of the workup
so that the true amount of remaining starting material
15 must be calculated by adding the amount 19 to the
amount 15.) Also, the ratio of 2-methylheptane (16) to
elimination products 17 and 18 was higher in the quartz
vessel, indicating that the used Pyrex flask affects the
formation of dehydrohalogenation products 17 and 18.
Note that the deuterium content of 16 is much lower for
the experiment carried out in the quartz vessel (21% D,
experiment 2) than that of 16 in the corresponding
experiment in used Pyrex (47% D, experiment 1). Fi-
nally, a much higher formation of dimers was noted in
quartz (6% dimers, experiment 2) than in used Pyrex
(trace dimers, experiment 1). Thus, once again, the data
show that a used Pyrex surface has some effect on the
reaction of alkyl iodides with LAD, although the major
pathways of the reaction appear to be the same in either
vessel. In comparison, very little difference between the
reactions carried out in used Pyrex (experiment 3) and
Teflon (experiment 4) at the lower stoichiometry of 1:0.2
was noted.
Sch em e 3
Effect of th e Lea vin g Gr ou p on th e Rea ction of
2-Ha lo-2-m eth ylh ep ta n es w ith LAD. The effect of the
halide leaving group was studied by varying the halide
from iodide to bromide to chloride. Since the reduction
potential becomes less favorable in the order alkyl iodides
> bromides > chlorides, the SET reaction rates of alkyl
halides would be expected to be in the same order.
If carbocation 23 is formed (step g), then it would be
expected to react much more rapidly with LiAlD4 (step
j) to form 16d 1 than with THF to form 16 (eq 6). J ust
the opposite is observed; 16 is formed in higher yield than
16d 1. Because deuterium incorporation is as low as 8%
(Table 3, experiment 2), step j would have to play, at best,
a minor role. Also, the formation of carbocation 23 (step
g) would be expected to be a slow reaction compared to
the electron-transfer process (step b), especially between
15 and LAD. A further argument against carbocation
23 acting as a major intermediate in the reaction of 15
with LAD lies in the results of the reaction of 25 (the
unsaturated counterpart of 15) with LAD to form radical
cyclized products (as discussed in the next section). The
comparable carbocation cyclizations would be expected
to be much slower than the corresponding radical cy-
clizations.9
III. Rea ction s of Un sa tu r a ted Ter tia r y Alk yl
Ha lid es w ith LAD. To further test the conclusions
formed in the reactions of the saturated halides 15, 20,
and 21, we decided to study the reactions of the compa-
rable terminally unsaturated halides.
Effect of Stoich iom etr y on th e Rea ction of 6-Iod o-
6-m eth yl-1-h ep ten e (25) w ith LAD. The tertiary
iodide 25 was allowed to react with a 5-fold excess of LAD
(eq 8 and Table 6, experiment 1). The reaction of 25 with
LAD produced almost exclusively the reduced cyclized
alkane 27 with no evidence of the reduced uncyclized
counterpart of 25. This observation argues against the
The expected trend of reaction rates was indeed
observed. Alkyl iodide 15 (Table 5, experiment 1) reacted
with LAD faster than did alkyl bromide 20 (experiment
2), which reacted faster than did alkyl chloride 21
(experiment 3). Even though the reaction rates varied
dramatically, the distribution of products remained
remarkably similar, indicating that electron transfer to
form the alkyl halide radical anion 22 (Scheme 3, step
b) is the rate-determining step of the reaction. The
product ratios of 16/17/18 were almost identical for the
alkyl iodide 15 and bromide 20. All three reactions
indicated significant amounts of hydrogen atom abstrac-
tion to form the nondeuterated alkane 16.
It is important to note that previous mechanistic
studies involving LAH or LAD support SET as a viable
reaction pathway for alkyl iodides and, to a lesser extent,
alkyl bromides. However, this is the first known report
of SET from LAD to an alkyl chloride. The reason for
observing SET with the tertiary chloride 21 while not
observing SET for primary and secondary alkyl chlorides
probably is due to the difference in the alkyl radical
stabilities.
P r op osed Mech a n ism of Red u ction of 2-Ha lo-2-
m eth ylh ep ta n es by LAD. As shown in Scheme 3, it is
(9) (a) Hill, E. A.; Davidson, J . A. J . Am. Chem. Soc. 1964, 86, 4663.
(b) Hill, E. A.; Theissen, R. J .; Taucher, K. J . Org. Chem. 1969, 34,
3061. (c) Hill, E. A.; Theissen, R. J .; Doughty, A.; Miller, R. J . Org.
Chem. 1969, 34, 3681. (d) Kossa, W. C., J r.; Rees, T. C.; Richey, H. G.,
J r. Tetrahedron Lett. 1971, 37, 3455.
(8) Bingham, R. C.; Schleyer, P. v. R. J . Am. Chem. Soc. 1971, 93,
3189.

Evidence for a SET Pathway
J . Org. Chem., Vol. 63, No. 22, 1998 7711
Ta ble 4. Effect of Rea ctor Su r fa ce on th e Rea ction of 2-Iod o-2-m eth ylh ep ta n e (15) w ith LAD in THF a
exp
1
rxn vessel
RI/LAD
1:5
time, h
% 15
% 16 (% D)
% 17
% 18
% 19
% dimers
material balance
used Pyrex
0.5
1
2
47
57
33
8
trace
61
43
27
5
2
5
5
7
16
20
35
51
61
10
13
17
22
23
23
4
5
8
20
27
6
30
13
15
9
trace
7
8
5
4
2
1
8
5
6
5
3
15
5
11
8
100
98
97
95
99
99
95
93
96
96
99
95
87
87
91
94
95
90
90
86
3 (62)
7 (59)
17 (51)
26 (47)
11 (16)
17 (19)
25 (19)
40 (21)
45 (21)
44 (21)
2 (14)
11 (10)
15 (8)
15 (8)
16 (8)
3 (13)
10 (10)
13 (9)
15 (9)
5
10
12
9
12
16
20
21
25
2
6
7
8
9
24
0.5
1
2
5
8
24
1
5
8
24
48
1
5
8
trace
1
2
3
5
5
6
trace
trace
2
3
2
3
quartz
1:5
trace
0
used Pyrex
Teflonb
1:0.2
1:0.2
81
60
49
40
35
68
61
48
37
4
4
3
6
7
9
trace
6
9
15
2
2
2
24
a
All reactions were carried out at a RI concentration of 0.070 M in an Ar atmosphere glovebox in the absence of light at room temperature.
b
Experiments 1, 3, and 4 employed Teflon-coated stir bars while experiment 2 was carried out in the absence of Teflon. Teflon was found
to be reactive toward LiAlH4, and thus, it is inappropriate for mechanistic studies involving LAH or LAD.7
Ta ble 5. Effect of Lea vin g Gr ou p on th e Rea ction s of 2-Ha lo-2-m eth ylh ep ta n es w ith LAD in THF a
exp
1
X
time
0.5 h
1 h
2 h
5 h
24 h
5 h
24 h
3 days
7 days
13 days
24 h
7 days
13 days
% RX
% 16 (% D)
% 17
% 18
% 19
material balance
I (15)
47
57
33
8
trace
89
66
26
4
2
5
5
7
16
20
35
51
61
9
23
43
56
59
trace
4
30
13
15
9
trace
2
1
2
0
0
100
98
97
95
99
100
100
95
93
96
3 (62)
7 (59)
17 (51)
26 (47)
trace
6 (82)
16 (73)
23 (73)
26 (69)
trace
10
12
trace
4
2
3
Br (20)
Cl (21)
8
10
11
trace
trace
2
0
92
92
77
0
0
0
92
100
96
4 (61)
8 (56)
9
a
Reactions were carried out at a RX concentration of 0.070 M at a RI/Nuc ratio of 1:5 in an Ar atmosphere glovebox in the absence of
light at room temperature in used Pyrex flasks and with the use of Teflon-coated stir bars.
Ta ble 6. Effect of Stoich iom etr y on th e Rea ction of
6-Iod o-6-m eth yl-1-h ep ten e (25) w ith LADa
When the stoichiometry of 25/LAD was reduced from
1:5 to 1:1, similar results were obtained (Table 6, experi-
ment 2) except that compound 27 contained a much lower
percent deuterium. This result was expected because
much less LAD was available to trap the intermediate
radical (as compared to experiment 1), and hence, more
radical was trapped by THF.
time,
h
%
% 27
%
material
exp RI/LAD
26 (% D) % 28 % 29 dimers balance
1
1:5
1
2
5
2
5
20 74 (79) trace trace trace
94
93
98
97
96
97
3
0
90 (78) trace trace
96 (78) trace trace
1
2
4
5
6
2
1:1
38 55 (27) trace trace
The reaction of the unsaturated tertiary iodide 25 with
LAD (Table 6, experiment 1) was found to be different
from the reaction of the corresponding saturated tertiary
iodide 15 (Table 5, experiment 1) in several ways. In the
saturated system, the products observed were the satu-
rated hydrocarbon 16 (and 16d 1), the alkenes 17 and 18,
and the alcohol 19 which was formed from iodide 15
during the workup procedure (eq 5). For the unsaturated
tertiary system, none of the uncyclized hydrocarbon
6-methyl-1-heptene (corresponding to alkane 16 in the
saturated tertiary system) was formed, and the dienes
28 and 29 (corresponding to alkenes 17 and 18) were
formed only in trace quantities. The low yield of dienes
in the reaction of alkyl iodide 25 with LAD was surprising
because the elimination products 17 and 18 represented
the major products in the saturated tertiary system.
6
0
85 (22) trace trace
91 (21) trace trace
11
a
Reactions were carried out at a RI concentration of 0.070 M
in an Ar atmosphere glovebox in the absence of light at room
temperature in used Pyrex flasks and with the use of Teflon-coated
stir bars.
-
formation of a carbocation because AlD4 would be
expected to trap the carbocation rapidly.
The results of the reaction of the cyclizable tertiary
radical probe 25 were also surprising in light of the
results of the radical probe studies of cyclizable primary

7712 J . Org. Chem., Vol. 63, No. 22, 1998
Ashby and Welder
and secondary systems. For example, when the hindered
primary alkyl iodide 1 was allowed to react with LAD,
the uncyclized alkene 2 was formed in 8% yield (as shown
in eq 1 for LAH). We reported in 198310 that the
secondary iodide, 6-iodo-1-heptene, was found to yield
1-heptene in 42% yield upon reaction with LAD. How-
ever, in the tertiary system, as reported in Table 6, the
uncyclized alkene, 6-methyl-1-heptene, was not observed.
Because the product distribution in the reaction of the
unsaturated tertiary alkyl iodide 25 with LAD was not
as expected, the reaction was carefully monitored.
Careful monitoring of the reaction of alkyl iodide 25
with a 2-fold excess of LAD revealed the rapid formation
of cyclized iodide 26 as a major intermediate in the
reaction (eq 9).11 (Cyclized iodide 26 was isolated from
It is interesting to note the difference in the reactivity
of 1-(iodomethyl)-3,3-dimethylcyclopentane (9) and alkyl
iodide 26. As discussed previously,2 when cyclized iodide
9 was allowed to react with LAD, 1,1,3-trimethylcyclo-
pentane (3) was formed in 99% yield containing 98% D.
On the other hand, significant hydrogen atom abstraction
was observed in the reaction of alkyl iodide 26 with LAD
as noted by the low deuterium incorporation in 1,1,2-
trimethylcyclopentane (27). Therefore, it is suggested
that cyclized iodide 26 is sufficiently hindered as to give
significant evidence of SET upon reaction with LAD
whereas alkyl iodide 9 is not sufficiently hindered.
The reaction of alkyl iodide 25 with LAD showed
similar results whether the reaction was carried out in
a used Pyrex or a quartz vessel.
Effect of th e Lea vin g Gr ou p on th e Rea ction s of
th e 6-Ha lo-6-m eth yl-1-h ep ten es w ith LAD. As shown
in eq 9, the unsaturated tertiary iodide 25 first cyclized
to 26, which was then reduced to the cyclized alkane 27.
By comparison, very little of the corresponding alkyl
chloride 33 reacted over a 14 day period (eq 10).
the reaction and analyzed by 1H NMR and GC/MS to
confirm its structure.) It is apparent from the data that
the tertiary iodide 25 rapidly undergoes a halogen-atom
radical chain process to form the primary iodide 26,
which is then reduced by LAD to alkane 27. Therefore,
this reaction does not accurately represent the reduction
of a tertiary alkyl iodide but instead represents the
reduction of a hindered primary alkyl iodide (26).
Even though only trace quantities of reduction prod-
ucts 34 and 27 were observed, the deuterium content of
these products was determined. The deuterium incor-
poration in the uncyclized alkane 34 was quantitative
(99% D) whereas the deuterium incorporation in the
cyclized alkane 27 was quite low (23% D), indicating
significant protium incorporation in the cyclized product
which suggests a radical precursor.
Sch em e 4
P r op osed Mech a n ism of Red u ction of 6-Iod o-6-
m eth yl-1-h exen e (25) w ith LAD. The proposed mech-
anism of reaction of iodide 25 with LAD is shown in
Scheme 5. It is suggested that iodide 25 can react by an
E2 mechanism (step a) to yield dienes 28 and 29. The
SET process (step b) is initiated by an electron transfer
from LAD to the alkyl iodide 25, resulting in the radical
anion 30 which rapidly dissociates (step c) to radical 31
and further cyclizes to radical 32 (step d) from which
both dimers and compounds 26, 27, and 27d 1 are formed.
It is suggested that cyclized iodide 26, a hindered primary
alkyl iodide, reacts with LAD in an SN2 or a SET fashion.
An SN2 reaction (step k) would result in the deuterated
alkane 27d 1 whereas a SET reaction would lead to radical
anion 35 (step l) which can lose an iodide ion to form
radical 32 (step m). It is clear from the data that the
major reaction for the formation of 27 is by the SET
process (Table 6, experiment 2). Radical 32 then follows
one of the available pathways (e, f, h, or i) until a stable
product is formed. Another available pathway for alkyl
iodide 25 is the E1 pathway that leads to the carbocation
36 (step n). Carbocation 36 either can react with LAD
or its byproducts in a SET fashion (step o) to form radical
31 (which then reacts as discussed above) or can yield
elimination products 28 and 29 (step p). Because no
uncyclized alkene was observed from alkyl iodide 25, the
As shown in Scheme 4, the formation of the cyclized
iodide 26 is similar to the mechanism reported earlier
by us1a,c and others.12 Because no uncyclized 6-methyl-
1-hexene was observed, it is proposed that cyclization of
31 to 32 is faster than hydrogen- or deuterium-atom
abstraction to form the uncyclized alkene.
(10) Ashby, E. C.; DePriest, R. N.; Pham, T. N. Tetrahedron Lett.
1983, 24, 2825.
(11) No internal standard for GC analysis was added so that
isolation of the cyclized iodide 26 would be more convenient. Therefore,
the relative areas approximate the molar ratios of the components
shown. This is the ONLY mechanistic experiment reported in which a
GC internal standard was not used.
(12) Brace, N. O. J . Org. Chem. 1967, 32, 2711.

Evidence for a SET Pathway
J . Org. Chem., Vol. 63, No. 22, 1998 7713
Sch em e 5
2-Br om oocta n e (12). Alkyl bromide 12 was synthesized
from 2-octanol in a manner analogous to that reported for
3-bromo-1-phenylprop-1-ene.13 The product was purified by
silica chromatography using hexane as the eluent. 1H NMR
(CDCl3): δ 4.20-4.07 (m, 1H), 1.94-1.73 (m, 2H), 1.71 (d, 3H,
J ) 6.7 Hz), 1.55-1.22 (m, 8H), 0.89 (m, 3H).
2-Meth yl-2-h ep ta n ol (19). Alcohol 19 was synthesized via
a Grignard reaction from 1-bromopentane and acetone in a
manner analogous to that reported for 2-methylhexan-2-ol.14
Distillation of the crude product (64.0-66.0 °C/12 mmHg)
followed by silica column chromatography (petroleum ether
to petroleum ether/diethyl ether mixtures as eluent) afforded
the purified alcohol in 63% yield. 1H NMR (CDCl3): δ 1.51-
1.41 (m, 2H), 1.41-1.23 (m, 7H), 1.21 (s, 6H), 0.90 (t, 3H, J )
6.9 Hz). MS m/e (relative intensity, C8H18O): 115 (80), 97 (23),
59 (100), 55 (44), 43 (51).
2-Ch lor o-2-m eth ylh ep ta n e (21). Alkyl chloride 21 was
synthesized from 2-methyl-2-heptanol and HCl in a manner
analogous to that reported for tert-butyl chloride.15 The crude
product was purified by silica chromatography with hexane
as eluent. 1H NMR (CDCl3): δ 1.79-1.70 (m, 2H), 1.57 (s,
6H), 1.55-1.22 (m, 6H), 0.91 (t, 3H, J ) 6.9 Hz). MS m/e
(relative intensity, C8H17Cl): 135 (0.2), 133 (0.7), 121 (0.2), 69
(54), 57 (75), 56 (100), 55 (57), 43 (53), 41 (71).
2-Br om o-2-m eth ylh eptan e (20). 2-Methyl-2-heptanol (2.52
g, 19.4 mmol) was added to a flask followed by concentrated
hydrobromic acid (2.5 mL). The reaction mixture was stirred
for 18 h at room temperature. The resulting organic layer was
washed with water until neutral to litmus and then dried over
calcium chloride. The solvent was removed by rotary evapora-
tion, and distillation of the residue (70.8-72.0 °C/18 mmHg)
afforded the desired product (1.86 g, 49.7%). Note that tertiary
bromide is not stable on silica or neutral alumina or under
the thermal conditions required for preparative GC. 1H NMR
(CDCl3): δ 1.81-1.73 (m, 2H), 1.75 (s, 6H), 1.60-1.42 (m, 2H),
1.41-1.23 (m, 4H), 0.91 (t, 3H, J ) 6.9 Hz). MS m/e (relative
intensity, C8H18Br): 137 (0.8), 135 (0.8), 113 (90), 71 (95), 57
(100), 43 (86), 41 (67).
SN1 pathway is not a likely pathway for the reduction of
iodide 25 by LAD.
Su m m a r y a n d Con clu sion s
Because it has been convincingly shown that both
hindered2 and unhindered3 primary alkyl iodides and
secondary1a,c,10 alkyl iodides react with LiAlH4 via a
predominantly SET pathway, it is not surprising that
tertiary alkyl iodides would do likewise because tertiary
alkyl iodides should be more easily reduced to the
corresponding radical than primary or secondary alkyl
iodides. By choosing saturated alkyl halides for this
study, we eliminate the possibility of the results being
explainable by means of a halogen-atom radical chain
process as has been suggested to explain previous results
concerning the use of unsaturated alkyl halides as
cyclizable radical probes. The difference between tertiary
alkyl halides and primary and secondary alkyl halides
is that the tertiary alkyl halides have a better possibility
of reacting via a carbocation as well as a radical. In
addition to not expecting tertiary alkyl halides to ionize
readily in THF, convincing evidence has been presented
that supports the contention that the results reported
herein are not consistent with the major reaction path-
way proceeding via a carbocation.
Exp er im en ta l Section
Ma ter ia ls. Carbon disulfide was purchased from Fisher
Scientific and distilled from calcium chloride prior to use.
Diphosphorus tetraiodide (Aldrich) and concentrated hydro-
bromic acid (Baker) were used as received. 2-Iodooctane
(Lancaster) was distilled prior to use. Authentic samples of
2-methylheptane (Aldrich) and 1,1,3-trimethylcyclopentane
(Wiley Organics) were obtained in order to verify their forma-
tion in mechanistic studies. Additional information on materi-
als, instrumentation, deuterium content calculations, and
titrations of LAH/LAD solutions was reported previously.2
Red u ction of Alk yl Ha lid es. The procedure for the
reduction of alkyl halides by LAH or LAD has been previously
reported.2
(13) Vogel, A. I. Vogel’s Textbook of Practical Organic Chemistry,
5th ed.; J ohn Wiley & Sons: New York, 1989; p 565.
(14) Vogel, A. I. Vogel’s Textbook of Practical Organic Chemistry,
5th ed.; J ohn Wiley & Sons: New York, 1989; p 538.
(15) Vogel, A. I. Vogel’s Textbook of Practical Organic Chemistry,
5th ed.; J ohn Wiley & Sons: New York, 1989; p 556.

7714 J . Org. Chem., Vol. 63, No. 22, 1998
Ashby and Welder
2-Iod o-2-m eth ylh ep ta n e (15). The title compound was
prepared by a known method;16 however, a more detailed
procedure is reported in the Supporting Information. 1H NMR
(CDCl3): δ 1.92 (s, 6H), 1.65-1.20 (m, 8H), 0.91 (t, 3H, J )
6.6 Hz).
2-Meth yl-6-h ep ten -2-ol. The title compound was synthe-
sized in a manner analogous to that reported above for 19
(substituting 5-bromo-1-pentene for the 1-bromopentane) in
55% yield. 1H NMR (CDCl3): δ 5.89-5.76 (m, 1H), 5.10-4.97
(m, 2H), 2.12-2.03 (m, 2H), 1.50-1.42 (m, 4H), 1.30-1.20 (m,
7H). MS m/e (relative intensity, C8H16O): 113 (11), 95 (14),
59 (100), 43 (35).
6-Ch lor o-6-m eth yl-1-h ep ten e (33). Alkyl chloride 33 was
synthesized from 2-methyl-6-hepten-2-ol and HCl in a manner
analogous to that reported for tert-butyl chloride.15 1H NMR
(CDCl3): δ 5.90-5.73 (m, 1H), 5.07-4.96 (m, 2H), 2.13-2.04
(m, 2H), 1.80-1.72 (m, 2H), 1.67-1.53 (m, 2H), 1.58 (s, 6H).
MS m/e (relative intensity, C8H15Cl): 148 (3), 146 (10), 111
(19), 95 (80), 69 (100), 56 (66), 55 (53), 41 (86).
6-Iod o-6-m eth yl-1-h ep ten e (25). The title compound was
synthesized in a manner analogous to that reported for 15
(substituting 2-methyl-6-hepten-2-ol for 19). 1H NMR
(CDCl3): δ 5.90-5.75 (m, 1H), 5.10-4.96 (m, 2H), 2.16-2.08
(m, 2H), 1.93 (s, 6H), 1.65-1.60 (m, 4H). 13C NMR (CDCl3):
δ 138.8, 115.3, 52.5, 50.0, 38.2, 33.5, 27.8. MS m/e (relative
intensity, C8H15I): 169 (0.7), 128 (4), 127 (7), 112 (12), 111 (97),
69 (100), 57 (11), 56 (11), 55 (61), 53 (10), 43 (16), 41 (53), 39
(26).
Rea r r a n gem en t of 6-Iod o-6-m eth yl-1-h ep ten e (25) to
1-(Iod om eth yl)-2,2-d im eth ylcyclop en ta n e (26). To 6-iodo-
6-methyl-1-heptene (0.038 g, 0.16 mmol) was added THF (1.93
mL) followed by a homogeneous solution of LAD in THF (0.35
mL, 0.90 M, 0.32 mmol). After 30 min, the solution was
quenched by addition to cold water (1 mL), and the resulting
mixture was then extracted with diethyl ether (2 × 1 mL).
The combined organic layers were dried over anhydrous
MgSO4, filtered, and concentrated. GC indicated that the
desired compound was formed in 89% relative yield. (For this
1
experiment only, the absolute yield was not determined.) H
NMR (CDCl3): δ 3.23 (dd, 1H, J ) 9.3, 3.9 Hz), 2.94 (dd, 1H,
J ) 9.3, 11.4 Hz), 2.20-2.03 (m, 1H), 1.70-1.20 (m, 6H), 1.03
(s, 3H), 0.77 (s, 3H). MS m/e (relative intensity, C8H15I): 238
(0.9), 223 (1.5), 111 (100), 69 (65).
Ack n ow led gm en t. We are indebted to Sarah J .
Shealy and David E. Bostwick for their mass spectrom-
etry assistance. This work was supported by the Na-
tional Science Foundation (Grant CHE-8914309) and
the Georgia Institute of Technology, School of Chemistry
and Biochemistry.
Su p p or tin g In for m a tion Ava ila ble: Methods and ex-
perimental data for 15 (1 page). This material is contained
in libraries on microfiche, immediately follows this article in
the microfilm version of this journal, and can be ordered from
the ACS; see any current masthead page for ordering
information.
(16) Lauwers, M.; Regnier, B.; Van Eenoo, M.; Denis, J . N.; Krief,
A. Tetrahedron Lett. 1979, 20, 1801.
J O9807058