On the reduction of tertiary radicals by samarium diiodide (Sml2)

Article abstract of DOI:10.1139/v00-011

Reduction of o-iodopbeayl 3-methylbut-2-enyl ether with samarium diiodide generates mixtures of 3-isopropyl-2,3-dihydrobenzofuran and 3-(2-propenyl)-2-3-dihydrobenzofuran along with a small amount of dimer. If a source of deuterium is present during the r

Full text of DOI:10.1139/v00-011

791
On the reduction of tertiary radicals by samarium
diiodide (SmI2)
Tadamichi Nagashima, Alexey Rivkin, and Dennis P. Curran
Abstract: Reduction of o-iodophenyl 3-methylbut-2-enyl ether with samarium diiodide generates mixtures of 3-
isopropyl-2,3-dihydrobenzofuran and 3-(2-propenyl)-2,3-dihydrobenzofuran along with a small amount of dimer. If a
source of deuterium is present during the reduction, then the 3-isopropyl product predominates and this product is la-
beled with one deuterium. However, attempts to quench the putative tertiary organosamarium reagent by adding a deu-
terium source after the reduction were not very successful at room temperature. But at 0°C, the organosamarium
reagent was generated (at least to the extent of about 50%, as measured by deuterium quenching), and its decomposi-
tion was followed over time by a series of quenching experiments. The results suggest that tertiary radicals are reduced
to a significant extent by SmI2 to form an anionic (presumably alkylsamarium) species. This species is thermally unsta-
ble and decomposes to the corresponding reduced and eliminated products. The reduced product is consistently formed
in slight excess over the eliminated one, and the mechanism of formation of these products is not yet clear.
Key words: samarium diiodide, SmI2, reduction, alkyl samarium.
Résumé : La réduction de l’oxyde d’o-iodophényle et de 3-méthylbut-2-ényle par le diiodure de samarium conduit à un
mélange de 3-isopropyl-2,3-dihydrobenzofurane et de 3-(prop-2-ényl)-2,3-dihydrobenzofurane en plus de faibles
quantités de dimères. Si la réduction se fait en présence d’une source de deutérium. Le produit 3-isopropyle est alors
prédominant et il est marqué par un deutérium. Toutefois, tous les essais effectués à la température ambiante en vue de
piéger le réactif organosamarium tertiaire putatif par addition d’une source de deutérium après la réduction ont été
infructueux. Toutefois, à 0°C, on a généré le réactif organosamarium (au moins jusqu’à un niveau de 50% si l’on se
base sur le produit piégé au deutérium) et on a étudié sa décomposition en fonction du temps en faisant appel à des
expériences de piégeage. Les résultats suggèrent que le SmI2 provoque une importante réduction des radicaux tertiaires
avec formation d’espèces anioniques (probablement de l’alkylsamarium). Cette espèce est thermiquement instable et
elle se décompose pour donner les produits correspondants de réduction et d’élimination. Il y a toujours formation
d’une quantité légèrement supérieure de produit réduit par rapport au produit d’élimination et le mécanisme de forma-
tion de ces produits n’est toujours pas clair.
Mots clés : diiodure de samarium, SmI2, réduction, alkylsamarium.
Nagashima et al. 799
Introduction
4). Little is known about the structure of these intermediates,
and they sometimes have limited stability (4a, 5). But they
often behave like “carbanions” and combine with reactive
electrophiles to give trapped products (1–4).
Samarium diiodide (SmI2) is a powerful and soluble re-
ducing agent whose popularity in organic synthesis is con-
tinually increasing (1). The use of samarium diiodide to
control sophisticated sequences of radical and ionic reac-
tions (2) rests on the understanding of how SmI2 reacts with
functional groups, radicals, and radical ions. It is well dem-
onstrated that the reductions of halides by SmI2 initially pro-
duce radicals (1), and it is now generally accepted that these
radicals often can be reduced by another equivalent of SmI2
to give putative organosamarium intermediates (eq. [1]) (3,
Reduction by SmI2 occurs readily for radicals bearing
electron withdrawing groups as well as primary alkyl radi-
cals. In contrast, most vinyl and aryl radicals appear to un-
dergo radical–solvent reactions (H-abstraction, addition)
faster than they are reduced by SmI2 (however, when elec-
tron withdrawing groups are present, vinyl radicals may be
reduced to vinylsamarium reagents. See ref. (6)). The situa-
tion with secondary and especially tertiary radicals is less
clear. Labeling studies (4a, b) with electrophiles show that
secondary radicals can be reduced to organosamarium re-
agents, but the trapped products are frequently accompanied
by small amounts of products of (apparent) radical
disproportionation. Tertiary radicals frequently do not give
trapped products at all (4–6). (Qualitatively similar behavior
is seen in solid phase reactions. See ref. (7)).
Received September 14, 1999. Published on the NRC
Research Press website on June 2, 2000.
We dedicate this paper to Prof. Stephen Hanessian on the
occasion of his 65th birthday.
T. Nagashima, A. Rivkin, and D.P. Curran.1 Department of
Chemistry, University of Pittsburgh, Pittsburgh PA 15260, U.S.A.
1Author to whom correspondence may be addressed.
Telephone: (412) 624-8240. Fax: (412) 624-9861.
e-mail: curran+@pitt.edu
Can. J. Chem. 78: 791–799 (2000)
© 2000 NRC Canada

792
Can. J. Chem. Vol. 78, 2000
Fig. 1. Generation and reaction of 3°-radicals by SmI2. (a) See Curran and Totleben (4b); (b) See Inanaga et al. (8); (c) See Molander
and McKie (4c).
Three key reactions of precursors of tertiary radicals in
the presence of SmI2 are shown in Fig. 1. We reported (4b)
that the reduction of aryl iodide 1 by SmI2 in THF/HMPA
without a proton source provided products 2 and 3 in a
60:40 ratio and concluded that these products probably re-
sulted from radical disproportionation. This result is typical
of other related systems (7). However, Inanaga (8) reported
that a very similar reduction of 4 with SmI2/HMPA in the
presence of tert-butanol provided 5 in 99% yield. In an
intramolecular example, Molander and McKie (4c)
reported that the reduction of 6 with SmI2/HMPA in the
presence of 1% Fe(DBM)3 gave the cyclized product 7 in
57% yield along with 30% yield of a 1:1 mixture of 8 and 9
(and a trace of 10). These divergent results are difficult to
compare directly due to the different reaction conditions
and because the presence of the ketone in 6 may alter the
reduction pathway (9), so we decided to undertake a series
of experiments designed to probe how tertiary radicals
© 2000 NRC Canada

Nagashima et al.
Table 1. Reduction of 1. Product study.
793
Entry
Sml2/1
Ratioa 2:3:11
Combined yieldb
1
2c
2.2
1.1
54 : 43 : 3
53 : 44 : 3
85%
79%
1
aBased on the H NMR spectrum of the crude misture.
bBased on the weight of the crude mixture.
c7% of the iodine 1 was recovered.
react with SmI2. We show that both in the presence and ab-
sence of proton sources, tertiary radicals can be reduced by
SmI2. These resulting anionic species are unstable and de-
compose to reduced and eliminated products.
excess of 2 over 3 can be accounted for by a small amount
of two electron reduction followed by protonation (proton
source unknown). If about 10% of 2 is formed by two elec-
tron reduction, then this could even account for the observa-
tion that 1.1 equiv. of SmI2 was not quite sufficient to
consume all of 1. However, the authentic product ratios of
radical–radical reactions are not known, and the results in
Table 1 might also be explained by reduction of the tertiary
radical to an organosamarium reagent, followed by decom-
position to the observed products with regeneration of
Sm(II). This decomposition must be quite rapid, otherwise
more starting material should have been recovered from ex-
periment 2. This possibility is not unreasonable, since many
organometallic species decompose by disproportionation and
coupling reactions that do not involve free radicals.
To investigate the effect of proton sources, we next con-
ducted the reduction of 1 in the presence of alcohols, and the
results of these experiments are reported in Table 2. Consis-
tent with the results of Inanaga (8), reduction of 1 with 2.2
equiv. SmI2 in the presence of tert-butanol now provided
only protonated product 2 in 79% yield (Table 2, entry 1).
Neither the alkene 3 nor the dimer 11 was detected. The use
of methanol-d4 provided mostly the reduced product 2 along
with a trace of the alkene 3 (ratio 97:3). The yield was
somewhat diminished (54%), but 2 contained 89% deute-
rium at the methine position based on GC–MS analysis (Ta-
ble 2, entry 2). In two related experiments, 1 was reduced in
the absence of methanol, and then methanol-d4 was added
(prior to oxygen quenching) after 30 s and 5 min, respec-
tively, (Table 2, entries 3 and 4). As expected, this gave re-
sults similar to entries 1 and 2 of Table 1, although the
yields were not quite as high (we did not analyze for the
dimer in these experiments). Interestingly, the reduced prod-
uct 2 contained 7–8% deuterium.
Results and discussion
We initially chose to repeat the reduction of 1. If the prod-
ucts from this reaction are formed exclusively by radical–
radical reactions, then 2 and 3 should be formed in a 50:50
(not 60:40) ratio, and there should also be some dimer 11
from radical recombination and perhaps the other possible
regioisomeric alkene (not shown) from disproportionation.
Formation of products by radical–radical reactions also only
requires 1 equiv. of SmI2, whereas electrophile-trapped prod-
ucts generally consume 2 equiv. of SmI2. It is known that 4
equiv. of HMPA are required to generate SmI2 with maxi-
mum reducing power (3, 10), and for consistency all reac-
tions in this work employed those conditions. In addition,
aqueous quenches were not used in any reactions, since it is
also known that SmI2 does not react rapidly with water and
that water increases the reduction potential of SmI2 (in other
words, addition of water tends to promote reductions rather
than stopping them (11)). Reactions were quenched by bub-
bling dry air through the mixture until the color of SmI2 had
dissipated (a few seconds). Data for the initial reductions of
1 are shown in Table 1.
Reduction of 1 with 2.2 equiv. of SmI2 (Table 1, entry 1)
gave results very similar to those reported (4b), except that
we were now able to carefully separate and characterize a
small amount of dimer 11 (as a meso–d,l-mixture). The
starting iodide 1 was completely consumed in this reaction,
even though the characteristic purple color of the SmI2 never
dissipated. Products 2, 3, and 11 were formed in 85% com-
bined (isolated) yield in a ratio of 53:44:3. Thus, dimer 11 is
formed, but the regioisomeric alkene of 3 (not shown) is not,
and there is somewhat more 2 than 3. The reaction was re-
peated with 1.1 equiv. of SmI2 (entry 2), and this time the
SmI2 rapidly decolorized. The same three products were iso-
lated in the same ratio in 79% yield. The starting iodide 1
was recovered in 7% yield, thus accounting for the margin-
ally lower yield.
We next ran a series of experiments at 0°C in an effort to
see if a more long-lived organosamarium species could be
generated. Blank experiments using a slight deficiency of
SmI2 (1.8 equiv.) showed that at 0°C the color dissipated
about 30–40 s after the end of the addition of the iodide.
Thus, we choose 1 min as the shortest quench to ensure that
the deuterium source was not introduced prior to the con-
sumption of the starting material. In a series of experiments,
the time between completion of addition of 1 and addition of
the quench was extended from 1 min to 180 min, as shown
These results seem reasonably consistent with formation
of products by radical reactions, if one hypothesizes that the
© 2000 NRC Canada

794
Can. J. Chem. Vol. 78, 2000
Table 2. Reduction of 1 with alcohol additives.
Entry
Additive
Ratio 2:3b
Combined yield
%D in 2
1
2
3
4
tBuOH
CD3OD
After 5 min, CD3OD
After 30 sec, CD3OD
2 only
97:3
56:44
58:42
79%
54%
52%
68%c
—
89a
8a
7b
1
aBased on the H NMR spectrum of the crude mixture.
bBased on the GC–MS of the crude mixture.
cBased on the weight of crude mixture.
Table 3. Reduction of 1 at 0°C with CD3OD quenching.
Entry
Rxn time (min)a
Ratio 2D:2H:3b
% Yield (2H+2D+3)d
1
2
3
4
5
6
7
8
9
0
1
2
4
5
10
20
30
40
50
60
180
91:06:03
59:27:14
52:30:18
45:33:22
41:35:24
35:37:28
27:43:30
21:48:31
17:51:32
14:53:33
11:54:35
00:58:42
91
91
91
91
90
90
89
90
88
89
90
89
10
11
12
aTime between the addition of 1 and CD3OD.
1
bBased on H NMR spectrum.
cThe CD3OD was present when 1 was added.
dCombined isolated yield.
in Table 3, entries 2–12. The Table shows the %D incorpo-
ration of 2D as well as the ratio of 2D:2H:3. Accepting that
product 2D is a marker for the formation of an anionic
(organosamarium) species, these data show that at least
about 60% of 1 is reduced to an anionic species in the ab-
sence of an added proton source. Over 3 h, the evolution of
this species is revealed by tracing the decrease in the yield
of 2D (to 0%) and the increases in the yield of both 2H and
3. Over this time, the combined isolated yield of the three
products remains constant (about 90%). Accordingly, the
initial 59% of anionic species (reflected by the amount of
2D in entry 2) decomposed over 3 h to 31% 2H and 28% 3.
periments with TEMPO in place of methanol, as shown in
Table 4. In addition to 2 and 3, these experiments also pro-
vided the TEMPO quenched product 12, although the effi-
ciency of the TEMPO quenching (21–23%) at 10 min is
considerably less than that of the MeOD quenching (35°C,
see Table 3, entry 6). The experiments also prove that reac-
tions were complete when the alcohol or TEMPO quench
was added. A separate experiment showed that addition of 1
and SmI2 to TEMPO provided recovered starting material 1;
presumably, the TEMPO was reduced by SmI2.
To complement the quenching studies with deuterated
methanol, we next prepared hexa-deutero substrates 13a, b
shown in Table 5 (see the Experimental section for their syn-
thesis). These substrates were reduced as usual with 1.1
We have recently shown that organosamarium reagents re-
act with TEMPO (12), so we tried duplicate quenching ex-
© 2000 NRC Canada

Nagashima et al.
795
Table 4. Reduction of 1 at 0°C with TEMPO quenching.
Run
Ratio 1:2:12a
Yield of 12b
1
2
45:32:23
49:30:21
16
19
1
aBased on the H NMR spectrum of the crude mixture.
bIsolated yield.
Table 5. Reduction of deuterated analogs of 1.
Entry
X
Ratio 14:15:16:17:13 (15:16)a
1
2
1
I
Br
16:44:34:04:02 (56:44)
48:29:23:02:20 (56:44)
aBased on H NMR spectrum of the product mixture.
equiv. of SmI2, and then the product ratios were assessed by
recording and integrating the crude H NMR spectra of the
3º-alkylsamarium reagent, then this could explain why 20%
of bromide remained after the samarium was consumed.
However, the bromide and iodide must yield the same radi-
cal after reduction, so it is unclear how the nature of the
original radical precursor can affect the fate of the cyclized
radical. Due caution should be exercised in interpreting
these experiments due to possible experimental error; in
each experiment, the recorded yield of the d7 product 15
somewhat exceeded that of the d5 alkene product 16. This is
not possible according to any mechanism, since the average
total of deuteria per molecule must be six.
1
reaction mixtures. The results of these two experiments are
shown in Table 5, entries 1,2. While the reduction of the io-
dide 13b was instantaneous, the bromide 13a required 2 h
for decolorization of the SmI2, and 20% unreacted bromide
was still recovered after workup. The formation of substan-
tial amounts of d7 product 15 in both cases clearly shows
that protonation cannot account for formation of all the re-
duced products and that some form of disproportionation is
occurring. Interestingly, the reduction of the (less reactive)
bromide 13b gave considerably more of the d6 (protonated)
product 14. If the increased amount of 14 in the case of bro-
mide was a result of more efficient reduction to the
Finally, to gauge the effect of an added ketone, we con-
ducted reduction of 13a in the presence of ketone 18
(Fig. 2). This experiment is complicated because the ketone
© 2000 NRC Canada

796
Can. J. Chem. Vol. 78, 2000
Fig. 2. Coreduction of 13a and a ketone.
Fig. 3. Pathways for product formation reduction of 3°-halides by SmI2.
can be reduced in competition with the iodide. As before,
disproportionation products 14/15/16 were produced and the
reduction product 14/15 contains a significant amount (63%)
of deuterium in the methine position. The ratio of reduced
product 14/15 to alkene 16 is also higher than in the experi-
ments without the ketone. As for the fate of the ketone 16 it-
self, 56% is recovered unchanged while 4% is trapped as the
THF adduct 19 (this product is thought to result from
H-abstraction from the THF followed by reduction to an
organosamarium reagent and addition to the ketone) (4).
About 27% of ketone 18 is reduced to alcohol 20. In turn,
this alcohol also contains 64% D, showing that there is a
significant pathway for transfer of allyl hydrogen from 13a.
to the radical 21 (that is, the reduction step is reversible). In
either case, an equivalent of Sm(II) must be regenerated in
tandem with disproportionation.
When the reduction is carried out in the presence of a pro-
ton source at either 25°C or 0°C, a good yield of the labeled,
reduced product 23 is formed. This strongly suggests that
the tertiary organosamarium species 22 is formed. Whether
or not the organosamarium species 22 is formed when the
reduction is conducted in the absence of proton source is
less clear. While proton sources are known to boost the re-
duction potential of SmI2 in the absence of HMPA (11), the
strong ligating capabilities of HMPA for Sm (10) suggest
that it would dampen or eliminate any effect of added proton
sources. If this is correct, then the reductions in the absence
of the proton source should also provide the tertiary alkyl sa-
marium reagent 22, and we interpret our inability to effi-
ciently trap this intermediate by subsequent addition of a
proton source as due to its instability with respect to
disproportionation. Supporting this analysis are the results of
the study at 0°C (Table 3), in which a significant amount of
the organosamarium reagent was generated and its decompo-
sition to the disproportionation products was followed over
time.
Conclusions
The results in this paper provide new insights into the be-
havior of tertiary radicals in the presence of SmI2. Some of
the possible reaction pathways are illustrated in Fig. 3. Ter-
tiary radical 21 could directly recombine and disproportion-
ate to give the observed products 24–26. Alternatively, it
could be reduced to a tertiary alkylsamarium species 22,
which in turn provides the final products either by
protonation (to give 23) in competition with bimolecular
disproportionation (to give 24–26) and (or) homolysis back
While the results support the postulate that a substantial
amount of the products derive from alkylsamarium reagents,
© 2000 NRC Canada

Nagashima et al.
797
they certainly do not exclude the formation of some prod-
ucts through radical reactions. For example, the dimer 24
may form through a radical recombination (although we can-
not exclude the possibility of coupling of organosamarium
reagents), and the inability to ever observe quantitative deu-
terium incorporation in the reduced product 23 may point to
the occurrence of radical hydrogen transfer reactions from
the solvent. In addition, the mechanism of the decomposition
of the intermediate organosamarium reagent 22 is also not
clear. This reagent could return to the radical 21 by the re-
verse of the reduction or it could decompose by a number of
ionic pathways. At this point, the radical pathway seems less
likely since the ratio of disproportionation products to re-
combination products is higher than that for typical tertiary
radicals.
Typical procedure for the reduction of O-prenyl-2-
iodophenol (1) in the presence of a proton source
HMPA (0.25 mL, 1.44 mmol) was added to SmI2 (0.10 M
t
in THF, 3.3 mL, 0.33 mmol) at 23°C. After 30 min, BuOH
(16 µL, 0.17 mmol) was added, and 5 min later, O-prenyl-2-
iodophenol (1) (42.7 mg, 0.148 mmol) in THF (0.5 mL) was
added in one portion at 23°C. After 30 min, dry air was bub-
bled for 1 min, and then 0.1 N HCl (10 mL) was added. The
mixture was extracted with Et2O (3 × 10 mL). The com-
bined ether layers were washed with 5% Na2S2O3 (1 × 10 mL),
sat. NH4Cl (1 × 15 mL), and brine (1 × 15 mL), and were
dried over MgSO4. The mixture was passed through a SiO2
column (eluent: hexanes/Et2O = 2:1) to give 3-(2-propyl)-
2,3-dihydrobenzofuran (2) (19.0 mg, 0.117 mmol, 79%
yield).
In summary, tertiary halides are not very cooperative part-
ners in a number of C—C bond forming processes mediated
by SmI2. In the past, this has been interpreted by us and oth-
ers as due to the inability of SmI2 to reduce 3°-radicals to
samarium reagents. Our new results suggest that SmI2 is
more efficient at reducing 3°-radicals than previously
thought, and that the problems with subsequent bond form-
ing reactions arise due to the instability of the samarium re-
agents, which disproportionate over a period of minutes or
less at room temperature.
General procedure for the reduction of O-prenyl-2-
iodophenol (1) by SmI2–HMPA with CD3OD quenching
To SmI2 (0.10 M in THF, 40.0 mL, 4.00 mmol) was added
HMPA (3.5 mL, 20 mmol) at 23°C. After 1 h, 3.82 mL of
the mixture was transferred to another flask, and cooled to
0°C. A solution of O-prenyl-2-iodophenol (1) (50 mg,
0.17 mmol) in THF (0.3 mL) was added dropwise over the
period of 20 s, and 5 min later, CD3OD (0.2 mL, 4.9 mmol)
was added. After 2 min, air was bubbled through the mixture
until the color turned to yellow. After the addition of 0.1 N
HCl (5 mL), the mixture was extracted with Et2O (3 ×
5 mL). The combined ether layers were concentrated and the
mixture was passed through a short SiO2 column (eluent: di-
chloromethane) to give a mixture of products (25.3 mg). The
Experimental
1
product ratio was determined by H NMR spectrum of the
mixture.
Typical procedure for the reduction of O-prenyl-2-
iodophenol (1) by SmI2–HMPA in the absence of
electrophile
Ethyl 3-trideuteriomethyl-4,4,4-trideuterio-2-butenoate
Triethyl phosphonoacetate (20.5 mL, 10.3 mmol) was
added dropwise into a suspension of NaH (2.15 g,
89.6 mmol) in THF 100 mL at 0°C. After 5 min, the cooling
bath was removed. After 1.5 h, acetone-d6 (8.0 mL, 0.11
mol) was added. After 2 h, the mixture was poured into 10%
NH4Cl (500 mL), and was extracted with Et2O (1 × 100 mL
and 2 × 50 mL). The combined ether layers were washed
with sat. NH4Cl (1 × 50 mL) and brine (1 × 50 mL), and
were dried over MgSO4. The product was purified by distil-
lation (bp = 76.0–76.5°C at 54 mmHg) to give the product
(7.57 g, 56.4 mmol, 62% based on NaH). IR (film)(cm–1):
HMPA (0.60 mL, 3.45 mmol) was added to SmI2 (0.10 M
in THF, 8.8 mL, 0.88 mmol) at 23°C. After 1 h, O-prenyl-2-
iodophenol 1 (0.116 g, 0.402 mmol) in THF (0.7 mL) was
added. After 21 h, dry air was bubbled through the mixture
until the color changed to light brown and then 0.1 N HCl
(20 mL) was added. The mixture was extracted with Et2O
(3 × 20 mL). The combined ether layers were washed with
5% Na2S2O3 (1 × 20 mL), sat. NH4Cl (1 × 20 mL), and
brine (1 × 20 mL), and were dried over MgSO4. The ratio of
the products (2, 3, and 11) was determined by recording the
1H NMR spectrum of the crude mixture. The products were
purified by flash chromatography (SiO2, hexanes/EtOAc =
15:1) to give the dimer 11 (1.6 mg, 2% yield) and a mixture
(34.4 mg) of 3-(2-propyl)-2,3-dihydrobenzofuran (1) and 3-
(propen-2-yl)-2,3-dihydrobenzofuran (2). The ratio of these
1
1694, 1144. H NMR (300 MHz, CDCl3) δ : 5.66 (1 H, s),
4.13 (2 H, q, J = 7.1 Hz), 1.26 (3 H, t, J = 7.1 Hz). 13C
NMR (75 MHz, CDCl3) δ : 166.21 (s), 155.72 (s), 115.95
(d), 58.97 (t), 26.0 (m), 18.8 (m), 13.97 (q). MS (EI) m/z:
134, 106, 89, 83. HRMS (EI) m/z: Anal. calcd. for
C7H6D6O2 134.1214; found: 134.1254.
1
products was determined by H NMR spectrum of the mix-
ture.
(dl,meso)-2,3-Bis(2,3-dihydrobenzofuran-3-yl)-2,3-
dimethylbutane (11)
3-Trideuteriomethyl-4,4,4-trideuterio-2-buten-1-ol
DIBALH (1.0 M in hexane, 100 mL, 0.10 mol) was added
dropwise to a solution of the above ester (6.76 g,
50.4 mmol) in CH2Cl2 (50 mL) at 0°C. After 1.5 h, MeOD
(6 mL) was added dropwise. After the mixture became cloudy,
sat. potassium sodium tartrate (25 mL) was added. The mix-
ture was poured into Et2O (150 mL), and 2 N HCl (150 mL)
was added. After the separation, the aqueous layer was ex-
tracted with Et2O (1 × 100 mL). The combined organic
1H NMR (300 MHz, CDCl3) (mixture of dl and meso) δ :
7.23 (2 H, d, J = 7.5 Hz), 7.19–7.09 (2 H, m), 6.91–6.77 (8
H, m), 4.78–4.66 (2 H, m), 4.42 (2 H, q, J = 9.4 Hz), 3.55 (1
H, dd, J = 8.6, 1.6 Hz), 3.48 (1 H, dd, J = 9.0, 3.1 Hz), 1.05
(3 H, s), 1.00 (3 H, s), 0.95 (3 H, s), 0.84 (3 H, s). MS
(EI) m/z: 322, 238, 161, 119. HRMS (EI) m/z: Anal. calcd.
for C22H26O2 322.1933; found: 322.1923.
© 2000 NRC Canada

798
Can. J. Chem. Vol. 78, 2000
layers were washed with 1 N HCl (1 × 50 mL) and brine (2
× 25 mL), and were dried over MgSO4. The product was pu-
rified by distillation at normal pressure (bp = 129–139°C) to
give alcohol (3.12 g, 33.9 mmol, 67%). IR (film) (cm–1):
3322, 1101, 1070. 1H NMR (300 MHz, CDCl3) δ: 5.41 (1 H,
t, J = 7.1 Hz), 4.20–4.08 (2 H, m), 1.20 (1 H, m). 13C NMR
(75 MHz, CDCl3) δ : 135.37 (s), 123.76 (d), 58.88 (t), 24.5
(m), 17.0 (m). MS (EI) m/z: 92, 74. HRMS (EI) m/z: Anal.
calcd. for C5H4D6O 92.1108; found: 92.1114.
O-(3-Trideuteriomethyl-4,4,4-trideuterio-2-buten-1-yl)-2-
bromo-phenol (13a)
Diethyl azodicarboxylate (2.96 g, 17.0 mmol) in THF
(2 mL) was added into a mixture of 2-bromophenol (2.78 g,
16.1 mmol), PPh3 (4.20 g, 16.0 mmol), and the above alco-
hol (1.38 g, 15.0 mmol) in THF (40 mL) at 0°C. The tem-
perature was gradually raised to 23°C. After 17.5 h, most of
the solvent was removed under reduced pressure, and
pentane (100 mL) was added to give a precipitate. After the
removal of the precipitate by filtration, the product was puri-
fied by flash chromatography (SiO2, hexanes/Et2O = 10:1)
followed by distillation (bp = 90–98°C at 0.2 mmHg) to give
13a (2.64 g, 10.7 mmol, 71% yield). IR (film) (cm–1): 1476,
O-(3-Trideuteriomethyl-4,4,4-trideuterio-2-buten-1-yl)-2-
iodophenol (13b)
Diethyl azodicarboxylate (2.96 g, 17.0 mmol) in THF
(2 mL) was added into a mixture of 2-iodophenol (3.52 g,
16.0 mmol), PPh3 (4.20 g, 16.0 mmol), and the above prenyl
alcohol-d6 (1.39 g, 15.1 mmol) in THF (35 mL) at 0°C. The
temperature was gradually raised to 23°C. After 16 h, most
of the solvent was removed under reduced pressure, and
pentane (100 mL) was added to give a precipitate. After the
removal of the precipitate by filtration, the product was puri-
fied by flash chromatography (SiO2, hexanes/Et2O = 30:1) to
give 13b (3.05 g, 10.4 mmol, 69% yield). IR (film) (cm–1):
1
1276, 1030, 744 H NMR (300 MHz, CDCl3) δ : 7.54 (1 H,
dd, J = 7.7, 1.6 Hz), 7.24 (1 H, dt, J = 1.6, 7.4 Hz), 6.90
(1 H, dd, J = 8.3, 1.1 Hz), 6.82 (1 H, dt, J = 1.5, 7.7 Hz),
5.51 (1 H, t, J = 6.6 Hz), 4.60 (2 H, d, J = 6.6 Hz). 13C
NMR (75 MHz, CDCl3) δ : 155.11, 137.21, 128.26, 121.60,
119.40, 113.52, 112.19, 65.93, (two CD3 carbons were not
detected). MS (EI) m/z: 248, 246, 228, 188, 186, 175, 173.
HRMS (EI) m/z: Anal. calcd. for C11H7D6BrO 246.0526;
found: 246.0490.
1
1469, 1275, 1239, 1016, 748. H NMR (300 MHz, CDCl3)
General procedure for the reduction of O-prenyl(d6)-2-
halophenol 13a, b by SmI2–HMPA
δ : 7.77 (1 H, dd, J = 7.7, 1.6 Hz), 7.28 (1 H, dt, J = 8.5,
1.6 Hz), 6.82 (1 H, dd, J = 8.2, 0.9 Hz), 6.70 (1 H, dt, J =
1.1, 7.6 Hz), 5.51 (1 H, t, J = 6.5 Hz), 4.58 (2 H, d, J =
6.5 Hz). 13C NMR (75 MHz, CDCl3) δ : 157.38, 139.68,
137.68, 129.30, 122.38, 119.52, 112.57, 86.89, 66.16, (two
CD3 carbons were not detected). MS (EI) m/z: 294, 221, 75.
HRMS (EI) m/z: Anal. calcd. for C11H7D6IO 294.0388;
found: 294.0382.
To SmI2 (0.10 M in THF, 11.0 mL, 1.10 mmol) was added
HMPA (0.85 mL, 4.89 mmol) at 23°C. After 20 min, O-
prenyl(d6)-2-iodophenol (13a) (0.299 g, 1.02 mmol) in THF
(1 mL) was added. After 8 h, dry air was bubbled until the
mixture turned to yellow, and then 0.1 N HCl (15 mL) was
added. After the extraction with Et2O (3 × 10 mL), the com-
bined ether layers were washed with 5% Na2S2O3 (1 ×
10 mL), 0.1 N HCl (1 × 10 mL), and brine (1 × 10 mL), and
were dried over MgSO4. The crude mixture was passed
through a SiO2 column (eluent: hexanes/Et2O = 10:1) to give
a mixture of products (0.130 g). The products ratio was de-
1-[2-(2,3-Dihydrobenzofuran-3-yl)propyl-2-oxy]-2,2,6,6-
tetra-methylpiperidine (12)
HMPA (4.0 mL, 23 mmol) was added to SmI2 (0.1 M in
THF, 44.0 mL, 4.40 mmol) at 23°C. After 14 min, the mix-
ture was cooled to 0°C, and O-prenyl-2-iodophenol (1)
(0.576 g, 2.00 mmol) in THF (2 mL) was added dropwise
over the period of 2 min. After 10 min, TEMPO (0.943 g,
6.04 mmol) in THF (1 mL) was added in one portion. After
1 min, a mixture of sat. potassium sodium tartrate (10 mL),
sat. NaHCO3 (5 mL), and 5% Na2S2O3 (5 mL) was added,
and the mixture was extracted with Et2O (3 × 15 mL). The
combined ether layers were washed with 5% Na2S2O3 (1 ×
15 mL), a mixture of sat. potassium sodium tartrate and sat.
NaHCO3 (5 mL each, 1 ×), H2O (2 × 15 mL), and brine (1 ×
15 mL), and were dried over MgSO4. The product was puri-
fied by flash chromatography (SiO2, hexanes/Et2O = 10:1) to
give 12 (0.122 g, 0.385 mmol, 19% yield). IR (film) (cm–1):
1483, 1460, 1375, 1364, 1234, 1132, 748.; 1H NMR
(300 MHz, CDCl3) δ : 7.33 (1 H, d, J = 7.3 Hz), 7.13 (1 H, t,
J = 7.4 Hz), 6.83 (1 H, t, J = 7.0 Hz), 6.78 (1 H, d, J =
8.0 Hz), 4.73 (1 H, dd, J = 9.5, 5.5 Hz), 4.54 (1 H, t, J =
9.6 Hz), 3.94 (1 H, dd, J = 9.5, 5.4 Hz), 1.68–1.45 (4 H, m),
1.38 (3 H, s), 1.37–1.26 (2 H, m), 1.16 (3 H, s), 1.154 (3
H, s), 1.146 (3 H, s), 1.12 (3 H, s), 1.11 (3 H, s). 13C NMR
(75 MHz, CDCl3) δ : 161.10 (s), 128.75 (s), 128.36 (d),
125.72 (d), 119.99 (d), 109.54 (d), 80.90 (s), 73.34 (t), 59.56
(s), 59.43 (s), 53.25 (d), 40.96 (t), 35.23 (q), 35.13 (q), 23.71
(q), 23.32 (q), 23.71 (q), 23.32 (q), 21.03 (q), 20.90 (q),
17.18 (t). MS (CI) m/z: 318 [M + H], 157, 142.
1
termined by H NMR spectrum of the mixture.
SmI2 reduction of O-prenyl(d6)-2-iodophenol (13a) in
the presence of the ketone 18
HMPA (0.95 mL, 5.5 mmol) was added to SmI2 (0.1 M in
THF, 11.0 mL, 1.10 mmol) at 23°C. After 50 min, a mixture
of O-prenyl(d6)-2-iodophenol (13a) (0.142 g, 0.483 mmol)
and 1,5-diphenylpentan-3-one (18) (0.120 g, 0.504 mmol) in
THF (1 mL) was added in one portion. After 1 h, dry air was
bubbled through the mixture, and then 0.1 N HCl (15 mL)
and brine (5 mL) were added. After the extraction with Et2O
(3 × 10 mL), the combined ether layers were washed with
5% Na2S2O3 (1 × 10 mL), 0.1 N HCl (1 × 10 mL), and brine
(1 × 10 mL), and were dried over MgSO4. The products
were separated by flash chromatography (SiO2, hex-
anes/EtOAc = 10:1 – 5:1) to give a mixture of 3-(2-propyl)-
2,3-dihydrobenzofuran (d6 and d7) (14 and 15) and 3-
(propen-2-yl)-2,3-dihydrobenzofuran (d5) (16) (61.1 mg), a
mixture of 1,5-diphenyl-3-pentanol (20) and 1,5-diphenyl-3-
(2-tetrahydrofuranyl)-3-pentanol (19) (39.1 mg), and 1,5-
diphenyl-3-pentanone (18) (67.0 mg, 56% recovery).
1,5-Diphenyl-3-(tetrahydrofuran-2-yl)pentan-3-ol (19)
The authentic sample of 19 was prepared as follows:
HMPA (0.95 mL, 5.46 mmol) was added to SmI2 (0.1 M in
© 2000 NRC Canada

Nagashima et al.
799
Chem. Ztg. 339, 101 (1997); (e) M. Murakami and Y. Ito. J.
Organometal. Chem. 473, 93 (1994); (f) G.A. Molander. In
Organic reactions. Vol. 46. Edited by L.A. Paquette. Wiley,
N.Y. 1994. p. 211. (g) G.A. Molander. Chem. Rev. 92, 29
(1992); (h) J.A. Soderquist. Aldrichimica Acta, 24, 15 (1991).
2. (a) G.A. Molander amd C.R. Harris. Tetrahedron, 54, 3321
(1998); (b) T. Skrydstrup. Angew. Chem. Int. Ed. Eng. 36, 345
(1997); (c) G.A. Molander and C.R. Harris. Chem. Rev. 96,
307 (1996); (d) D.P. Curran and R.L. Wolin. Synlett. 317
(1991); (e) T.L. Fevig, R.L. Elliott, and D.P. Curran. J. Am.
Chem. Soc. 110, 5064 (1988).
3. E. Hasegawa and D.P. Curran. Tetrahedron Lett. 34, 1717 (1993).
4. (a) D.P. Curran, T.L. Fevig, C.P. Jasperse, and M.J. Totleben.
Synlett. 943 (1992); (b) D.P. Curran and M.J. Totleben. J. Am.
Chem. Soc. 114, 6050 (1992); (c) G.A. Molander and J.A.
McKie. J. Org. Chem. 56, 4112 (1991).
5. G.A. Molander and J.A. McKie. J. Org. Chem. 58, 7216 (1993).
6. M. Hojo, H. Harada, J. Yoshizawa, and A. Hosomi. J. Org.
Chem. 58, 6541 (1993).
7. (a) X.H. Du and R.W. Armstrong. J. Org. Chem. 62, 5678
(1997); (b) X.H. Du and R.W. Armstrong. Tetrahedron Lett.
39, 2281 (1998).
8. J. Inanaga, O. Ujikawa, and M. Yamaguchi. Tetrahedron Lett.
32, 1737 (1991).
9. D.P. Curran, X. Gu, W. Zhang, and P. Dowd. Tetrahedron, 53,
9023 (1997).
THF, 11 mL, 1.1 mmol) at 23°C. After 40 min, a mixture of
iodobenzene (0.334 g, 1.64 mmol) and 1,5-diphenylpentan-
3-one (18) (113.9 mg, 0.478 mmol) in THF (1 mL) was
added. After 21 h, 0.1 N HCl (15 mL) was added. The
mixture was extracted with Et2O (3 × 10 mL). The com-
bined ether layers were washed with 5% Na2S2O3 (1 ×
10 mL), 0.1 N HCl (1 × 10 mL), and brine (1 × 10 mL), and
dried over MgSO4. The product was purified by flash chro-
matography (SiO2, hexanes/EtOAc = 5:1 – 3:1) to give 19
(0.124 g, 0.398 mmol, 83% yield). IR (film) (cm–1): 3461,
1495, 1453, 1066, 749, 699.; 1H NMR (300 MHz, CDCl3) δ :
7.37–7.13 (10 H, m), 3.94–3.77 (3 H, m), 2.83–2.57 (4
H, m), 2.08 (1 H, s), 2.07–1.82 (7 H, m), 1.73 (1 H, dt, J =
5.3, 12.6 Hz). 13C NMR (75 MHz, CDCl3) δ : 142.56 (s),
142.44 (s), 128.49 (d), 128.42 (d), 125.94 (d), 125.81 (d),
83.32 (d), 74.47 (s), 68.58 (t), 38.89 (t), 36.59 (t), 30.27 (t),
29.79 (t), 26.27 (t), 25.59 (t). MS (EI) m/z: 239, 91. HRMS
(EI) m/z: Anal. calcd. for C17H19O [M – C4H7O] 239.1436;
found: 239.1437.
Acknowledgments
We thank the National Institutes of Health for funding this
work.
10. (a) M. Shabangi and R.A. Flowers. Tetrahedron Lett. 38, 1137
(1997); (b) M. Shabangi, J.M. Sealy, J.R. Fuchs, and R.A.
Flowers. Tetrahedron Lett. 39, 4429 (1998).
11. E. Hasegawa and D.P. Curran. J. Org. Chem. 58, 5008 (1993).
12. T. Nagashima and D.P. Curran. Synlett. 330 (1996).
References
1. (a) A. Krief and A.M. Laval. Chem. Rev. 99, 745 (1999);
(b) R. Nomura and T. Endo. Chem. Eur. J. 4, 1605 (1998);
(c) G.A. Molander. In Comprehensive organic synthesis. Vol.
1. Edited by B.M. Trost and I. Fleming. Pergamon, Oxford.
1991. p. 251. (d) F.A. Khan and R. Zimmer. J. Prakt. Chem.
© 2000 NRC Canada