o-Bromo-p-methoxyphenyl ethers. Protecting/radical translocating (PRT) groups that generate radicals from C-H bonds β to oxygen atoms

Article abstract of DOI:10.1021/ja954150z

The o-bromo-p-methoxyphenyl ether group is introduced as a new protecting/radical translocating (PRT) group. This group protects an alcohol both before and after its use as a translocating group to generate a radical from a C-H bond β to the protected alc

Full text of DOI:10.1021/ja954150z

3
142
J. Am. Chem. Soc. 1996, 118, 3142-3147
o-Bromo-p-methoxyphenyl Ethers. Protecting/Radical
Translocating (PRT) Groups That Generate Radicals from C-H
Bonds â to Oxygen Atoms
Dennis P. Curran* and Jinyou Xu
Contribution from the Department of Chemistry, UniVersity of Pittsburgh,
Pittsburgh, PennsylVania 15260
X
ReceiVed December 11, 1995
Abstract: The o-bromo-p-methoxyphenyl ether group is introduced as a new protecting/radical translocating (PRT)
group. This group protects an alcohol both before and after its use as a translocating group to generate a radical
from a C-H bond â to the protected alcohol. All prior PRT groups generate radicals R to the functional group that
they protect. The group is introduced by Mitsunobu reaction or Williamson ether synthesis, and removed by oxidation
with ceric ammonium nitrate. The efficiency of the radical translocation reaction has been studied by isotopic labeling
experiments with tributyltin deuteride. These results were used to design and execute a series of tandem radical
translocation/cyclization reactions that illustrate the potential usefulness of the PRT group in synthesis. Secondary
radicals are generated with about 50% efficiency due to slow 1,5-hydrogen transfer and competing 1,6-hydrogen
transfer. Tertiary radicals are generated with efficiencies of about 80%. Modified PRT groups with added ortho
substituents (Br, Me) can increase the efficiency of radical generation up to about >80% and >90%, respectively.
The results provide only limited support for a rate and selectivity analogy based on radical cyclizations that was
used to design the groups.
Introduction
1
Protecting/radical translocating (PRT) groups are designed
to serve a dual function in the synthesis of target molecules:
they selectively activate a remote functionality in a molecule
for a radical bond-forming reaction and they serve as a
2
protecting group both before and after the radical reaction. Like
Figure 1.
3
standard protecting groups, they should be designed with ease
of introduction and removal in mind, and they should withstand
diverse sets of common reaction conditions. Beyond that, they
must function rapidly and selectively in the radical translocation,
the functional group that is protected. Figure 1 illustrates several
1
,5a,6a
5a,6b,8
PRT groups for alcohols,
amides/amines,
and
5
b,c,7
carboxylates.
The target hydrogen that is abstracted when
4
which is often an intramolecular hydrogen transfer reaction.
the PRT group is implemented in each structure is highlighted.
In all of the groups introduced to date, the target C-H bond is
R to the protected functional group. This location for the C-H
bond is a convenient one from the standpoint of the structure
of typical protecting groups, and it is also advantageous because
the functional group may weaken the target C-H bond, and
thereby facilitate radical translocation by hydrogen transfer.
Our recent detailed study of substituent effects on intra-
molecular 1,5-hydrogen transfer reactions in a simple model
At one key stage, the PRT group is implemented to activate
the C-H bond adjacent to the functional group, which then
allows a subsequent radical bond-forming reaction. The
synthesis then proceeds to the point where the PRT group is no
longer required for protection, at which point it is removed.
A number of PRT groups have been introduced recently from
1
,5
6
7,8
our lab and those of De Mesmaeker and others.
These
groups vary both in the nature of the protecting group and in
9
system provides guidelines for the design of PRT groups with
X
Abstract published in AdVance ACS Abstracts, March 15, 1996.
new features. In this study, we suggested that there was a
(1) Curran, D. P.; Kim, D.; Liu, H. T.; Shen, W. J. Am. Chem. Soc.
qualitative analogy between 5-exo-cyclizations and 1,5-hydrogen
1
988, 110, 5900.
2) For a synthetic application of the (o-bromophenyl)dimethylsilyl ether
group, see: Schwartz, C. E.; Curran, D. P. J. Am. Chem. Soc. 1990, 112,
272.
3) Greene, T. W.; Wuts, P. G. M. ProtectiVe Groups in Organic
Synthesis; Wiley: New York, 1991.
4) (a) Beckwith, A. L. J.; Ingold, K. U. In Rearrangements in the Ground
9
(
atom transfer reactions. In other words, radicals that suffered
rapid, selective 5-exo-cyclizations might also suffer rapid,
selective 1,5-hydrogen transfer reactions. This analogy suggests
that the importance of locating the target C-H bond R to the
functional group in a PRT substrate may be overrated, and that
the nature and geometry of the connecting chain between the
target C-H bond and the initial radical site on the PRT group
9
(
(
and Excited States; de Mayo, P., Ed.; Academic: New York, 1980; Vol. 1,
p 161. (b) Wilt, J. W. In Free Radicals; Kochi, J. K., Ed.; Wiley-
Interscience: New York, 1973; Vol. 1, p 333.
(5) (a) Snieckus, V.; Cuevas, J. C.; Sloan, C. P.; Liu, H.; Curran, D. P.
J. Am. Chem. Soc. 1990, 112, 896. (b) Curran, D. P.; Abraham, A. C.; Liu,
H. T. J. Org. Chem. 1991, 56, 4335. (c) Curran, D. P.; Liu, H. T. J. Chem.
Soc., Perkin Trans. 1 1994, 1377. (d) Curran, D. P.; Yu, H. S.; Liu, H.
T.Tetrahedron 1994, 50, 7343. (e) Yamazaki, N.; Eichenberger, E.; Curran,
D. P. Tetrahedron Lett. 1994, 35, 6623.
(7) (a) Esker, J. L.; Newcomb, M. Tetrahedron Lett. 1992, 33, 5913. (b)
Esker, J. L.; Newcomb, M. J. Org. Chem. 1994, 59, 2779.
(8) (a) Murakami, M.; Hayashi, M.; Ito, Y. J. Org. Chem. 1992, 57,
793. (b) Williams, L.; Booth, S. E.; Undheim, K. Tetrahedron 1994, 50,
13. (c) Undheim, K.; Williams, L. J. Chem. Soc., Chem. Commun. 1994,
883. (d) Brunckova, J.; Crich, D.; Yao, Q. W. Tetrahedron Lett. 1994, 35,
6619.
(
6) (a) Denenmark, D.; Hoffmann, P.; Winkler, T.; Waldner, A.; De
Mesmaeker, A. Synlett 1991, 621. (b) Denenmark, D.; Winkler, T.; Waldner,
A.; De Mesmaeker, A. Tetrahedron Lett. 1992, 33, 3613.
(9) Curran, D. P.; Shen, W. J. Am. Chem. Soc. 1993, 115, 6051.
0
002-7863/96/1518-3142$12.00/0 © 1996 American Chemical Society

o-Bromo-p-methoxyphenyl Ethers
J. Am. Chem. Soc., Vol. 118, No. 13, 1996 3143
yields by standard bromination reactions of the corresponding
1
4
phenol 10 with TBABr3 or Br2 (eq 1b). Four different PMP-
based groups were surveyed in this study: monobromides 9a
(R ) H), 9c (R ) Me), and 9d (R ) CMe2OTBS) and
dibromide 9b (R ) Br). The o-alkyl compounds 9c and 9d
were made from the dibromide 9b by halogen-metal exchange
and quenching, as shown in eq 1c.
Figure 2.
Figure 3.
may be more important in promoting rapid 1,5-hydrogen transfer
reactions than the bond dissociation energy of the target C-H
bond.
We envisioned that it would be useful to develop PRT groups
that could generate radicals â to protected functional groups.
Such groups could be used to abstract and subsequently
functionalize completely unactivated C-H bonds (that is, C-H
bonds not adjacent to any heteroatoms or π-conjugating groups)
through the intermediacy of simple alkyl radicals. Our design
of â-translocating groups was guided by the proposed analogy
between 5-exo-cyclizations and 1,5-hydrogen transfer reactions.
Cyclizations of aryl allyl ether radicals like 1 (Figure 2) are
exceptionally rapid and selective.10 This is presumably due to
the high reactivity of the aryl radical coupled with the favorable
geometry of the substrate for cyclization. The analogy then
suggests that 1,5-hydrogen transfer reactions of radicals like 3
might be sufficiently rapid and selective for use in PRT groups.
Beckwith has already observed that this type of 1,5-hydrogen
transfer was competitive with 6-exo-cyclization when R )
vinyl,1 and we set out to learn if these hydrogen transfers could
still occur when R ) alkyl.
Reductions of several substrates were initially conducted
under fixed concentration conditions with tributyltin deuteride
(0.01 M) to survey the efficiency of 1,5-hydrogen transfer
compared to 1,6-hydrogen transfer and direct reduction by
tributyltin deuteride. The results of these labeling studies are
summarized in eqs 2a-e. The products of the labeling
experiments were compared with authentic (fully protio) samples
to identify the structures and to locate the deuterium label. The
ratios listed in eqs 2a-e were determined by integration of the
2
H NMR spectrum of the crude reaction mixture. Isolated yields
were not determined in these reactions, but the suspicion that
they were high was confirmed by the subsequent preparative
experiments.
0
In this paper, we introduce the first class of PRT group that
generates radicals â to the protected functional group. The
p-methoxy-o-bromophenyl group11 protects alcohols, and it
generates secondary alkyl radicals from C-H bonds â to the
oxygen atom (see 6, Figure 3) with 50-55% efficiency and
tertiary alkyl radicals with 80-85% efficiency. This efficiency
can be increased to >80% for secondary radicals and >90%
for tertiary radicals by introducing a second o-bromine atom or
Reduction of the phenylethanol derivative 11 provided 12
containing 82% of the label in the benzylic position (1,5-
hydrogen transfer) and 18% of the label on the aromatic ring
(direct reduction). Phenylpropanol homolog 13 (eq 2b) provided
mainly the product 14 of 1,6-hydrogen transfer (72%) along
with smaller amounts of 1,5-hydrogen transfer (13%) and direct
reduction (15%) products. The reduction of cyclopentyl
substrate 15 (eq 2c) was especially unselective, providing
significant amounts of directly reduced product (35%) along
(better yet) an o-methyl group.
15
with the products of 1,4, 1,5 (two stereoisomers), and 1,6 (two
stereoisomers) hydrogen transfer. Unfortunately, the chemical
shifts of the hydrogens of cyclopentyl methylene groups were
too close to confidently assign these protons, so we could not
quantitate the ratio of 1,5/1,6-hydrogen transfer. However, we
suspect that the minor (2%) product should be the trans
Results and Discussion
We selected the o-bromo-p-methoxyphenyl PRT group to
integrate with the standard p-methoxyphenyl (PMP) ether
protecting group for alcohols.11 The PMP-based PRT groups
were introduced by standard O-alkylation reactions (either
Williamson ether synthesis12 or Mitsunobu reaction ) starting
from the appropriate o-bromophenol 8a or 8b and alcohol 7 or
halide (eq 1a; details are provided in the supporting information).
In turn, the o-bromophenols 8a,b were synthesized in good
13
16
stereoisomer resulting from 1,5-hydrogen transfer. Thus, even
if the major product (21%) is the cis isomer from 1,5-hydrogen
transfer, the total amount of 1,6-hydrogen transfer still exceeds
that of 1,5-transfer.
(
10) (a) Beckwith, A. L. J.; Gara, W. B. J. Chem. Soc., Perkin Trans. 2
(14) Shoji, K.; Takaaki, T.; Tsuyoshi, O.; Nakamura, H.; Masahiro, F.
Bull. Chem. Soc. Jpn. 1987, 60, 4187.
1
975, 795. (b) Abeywickrema, A. N.; Beckwith, A. L. J. J. Chem. Soc.,
Chem. Commun. 1986, 464. (c) Beckwith, A. L. J.; Meijs, G. F. J. Org.
Chem. 1987, 52, 1922. (d) Abeywickrema, A. N.; Beckwith, A. L. J.; Gerba,
S. J. Org. Chem. 1987, 52, 4072.
(15) 1,4-Hydrogen transfers are unusual but not unknown (see ref 4).
We cannot rule out the possiblity that part or all of the “1,4-hydrogen transfer
product” arises from bimolecular hydrogen transfer reactions.
(16) For the radical derived from 1,5-hydrogen transfer, we expect that
the deuterium atom will be delivered trans to the phenyl ether with a good
level of selectivity. For the radical derived from 1,6-hydrogen transfer, we
expect low stereoselectivity in the deuterium transfer reaction. See: Giese,
B. Angew. Chem., Int. Ed. Engl. 1989, 101, 993.
(
11) (a) Fukuyama, T.; Laird, A. A.; Hotchkiss, L. M. Tetrahedron Lett.
985, 26, 6291. (b) Petitou, M.; Duchaussoy, P.; Choay, J. Tetrahedron
Lett. 1988, 29, 1389.
1
(
(
12) Johnstone, R. A. W.; Rose, M. E. Tetrahedron 1979, 35, 2169.
13) Mitsunobu, O. Synthesis 1981, 1.

3
144 J. Am. Chem. Soc., Vol. 118, No. 13, 1996
Curran and Xu
radical, while the methyl group in 20 tips the balance in favor
of 1,5-hydrogen transfer by providing a tertiary radical. These
two effects appear to have different origins as judged by
comparing the total of 1,5- and 1,6-hydrogen transfer products
to the directly reduced product. In substrate 13, this total (85%)
is about the same as that for substrate 11 (82%). This implies
(
somewhat surprisingly) that the increase in the rate of 1,6-
hydrogen transfer of the radical derived from 13 is roughly
balanced by a decrease in the rate of 1,5-hydrogen transfer.
However, in substrate 20, the significant decrease in the yield
of the reduced product 21 implies that the methyl group
increases the rate of 1,5-hydrogen transfer. These results suggest
that useful radical translocation reactions based on the o-bromo-
p-methoxyphenyl group can be planned by using standard
substituent effects to favor 1,5-hydrogen transfer and by using
standard reaction techniques to minimize the tin hydride
concentration.
Equations 2d,e show the results of labeling experiments
coupled with subsequent cyclizations. Reduction of 17 (eq 2d)
under the standard conditions now provided two types of
products 18 and 19. The cyclized product 19 (54%) results
from 1,5-hydrogen transfer, and this was formed as an 8/1
mixture of cis/trans isomers that were labeled only in the methyl
group. The uncyclized product 18 resulted partly from direct
reduction (19%) and partly from 1,6-hydrogen transfer (27%).
Significantly better results were obtained with the methyl analog
To support these suggestions, we conducted a series of
preparative experiments with the o-bromo-p-methoxyphenyl
group. These experiments, along with all the subsequent
experiments with modified PRT groups, are summarized in
Table 1. In these reactions, tributyltin hydride and AIBN were
added by syringe pump to a solution of the precursor in refluxing
2
2
0 in eq 2e. This provided 93% of the 5-exo-cyclized product
2 again bearing the label in the methyl group (as a 1.2/1
mixture of isomers) along with 7% of the reduced product 21
17
benzene. After DBU workup, ratios of the reduced product
(
not shown) to the cyclized product were determined by GC
analysis. These products were separable in all cases, and Table
provides the isolated yields of purified cyclized products as
apparently resulting from 1,6-hydrogen transfer. Since reso-
2
nances in H NMR spectra are broad, such experiments could
1
fail to detect small amounts of minor products. Nonetheless, it
is clear that there is significantly less directly reduced product
in the reduction of 20 compared to 17.
The proposed mechanism for the reaction of 17 is illustrative
of all the substrates, and this is shown in eq 3. Bromine
abstraction to generate aryl radical 23 is followed by a
competition among direct reduction (leading to 24), 1,5-
hydrogen transfer (leading to 19 after 5-exo-cyclization of 27
and deuterium transfer), and 1,6-hydrogen transfer (leading to
6 after deuterium transfer). We conclude from the results in
eqs 2a-e that the hydrogen transfer reactions of aryl radicals
like 23 are sufficiently rapid to be useful in synthesis.
Intrinsic selectivity between 1,5- and 1,6-hydrogen transfer
does not appear to be high, as judged by the result of the
relatively unbiased substrate 17, which provided only a 2/1 ratio
of 19 (1,5-transfer product) to 26 (1,6-transfer product).
Fortunately, it appears that this bias can be shifted in either
direction by substituents. The phenyl group in 13 tips the
balance in favor of 1,6-hydrogen transfer by providing a benzyl
well as the cis/trans ratio.
Entries 1a and 2 show the preparative experiments that
correspond to the labeling experiments in eqs 2d and 2e. For
substrate 17a, pure product 29 was isolated in 44% yield, while
substrate 20 gave 30 in 75% yield. Each of these yields was
only slightly below the amount of cyclized product present as
judged by the cyclized/reduced (C/R) ratio. This shows that
total cyclized/reduced yields are high and that product loss
during purification is small. The ratios of cyclized/reduced
products are similar in the preparative and labeling experiments
2
(for 17a, 53/47 and 54/46, and for 20, 85/15 and 93/7). For
substrate 17a, this similarity is disappointing, and it indicates
that a significant amount of the reduced product (about 40%)
formed in the syringe pump experiment comes from direct
reduction of the aryl radical prior to any translocation; the
remainder of the reduced product comes from 1,6-hydrogen
transfer. The lower tin hydride concentrations used in the
(17) Curran, D. P.; Chang, C.-T. J. Org. Chem. 1989, 54, 3140.

o-Bromo-p-methoxyphenyl Ethers
J. Am. Chem. Soc., Vol. 118, No. 13, 1996 3145
Table 1. Radical Translocation/Cyclization Reactions
To briefly probe the effects of other substituents on the
radical, we selected the silyl ether and ketone substrates shown
in entries 5 and 6. Reduction of terminal alkene 35a provided
a surprisingly low yield of 36a (48%) and a correspondingly
low C/R ratio (56/44) when compared to the results of the
tertiary methyl system (entry 2, 75%, C/R ) 85/15). This led
us to suspect that a significant amount of the reduced product
resulted from failed cyclization. A reduced rate of cyclization
is not unreasonable because there is no Thorpe/Ingold effect in
this substrate and because related alkoxy-substituted radicals
1
9
are known to cyclize rather slowly. This explanation was
supported by reducing the more activated acceptor 35b to
provide an 84% isolated yield of 36b with a 94/6 C/R ratio. As
expected, the silyl ether is even better than a methyl group at
promoting 1,5-hydrogen transfer. Reduction of ketone 37 was
interesting, though not preparatively useful. In addition to the
formation of large amounts of reduced product (C/R ) 60/40),
this reaction produced the expected product of translocation/6-
endo-cyclization 38 in 12% yield and the unexpected spirocycle
3
9 in 27% yield. This last product results from an interesting
tandem cyclization where direct closure of the aryl radical to
20
the ketone precedes cyclization of the resulting alkoxy radical
2
1
to the alkene.
These results suggest that the parent o-bromo-p-methoxy-
phenyl ether PRT group will have moderate synthetic utility.
Though it may be useful for generating conjugated secondary
radicals (see eq 2a), this group does not appear to be useful for
generating simple secondary alkyl radicals. Maximum possible
yields of secondary radicals generated by 1,5-hydrogen transfer
(
8
as limited by 1,6-hydrogen transfer) are in the range of 70-
0%, but even these yields are experimentally difficult to obtain
due to direct reduction. Actual isolated yields in the range of
4
0-50% are obtained by standard experimental techniques. The
generation of tertiary alkyl radicals with this group is signifi-
cantly better. Maximum yields of tertiary radicals reach 80-
9
0%, and approaching these maxima is experimentally easier
due to the increased rate of 1,5-hydrogen transfer. Actual
isolated yields for tertiary substrates are in the 70-80% range.
To improve the 1,5-hydrogen transfer efficiency, we studied
several modified PRT groups. One advantage of the PRT
method is that the ultimate radical precursor (a C-H bond) is
not actually lost during the direct reduction. This means that a
given yield could be statistically increased by cleaving the
p-methoxyphenyl group from the reduced product and reintro-
ducing another o-bromo-p-methoxyphenyl group. However,
such recycling procedures are experimentally tedious. We
envisioned that a better alternative would be to use the “double-
a
OAr cis to H. b Relative configuration
The cis isomer has CH
2
c
d
determined by NOE. Relative configuration not assigned. Relative
configuration assigned by the γ-gauche effect in C NMR. Relative
configuration assigned by the analog to entry 5f. Relative configuration
assigned by desilylation; the minor isomer gives a lactone, whereas
the major isomer does not.
1
3
e
f
22
barreled” o,o-dibromo-p-methoxyphenyl group. If the directly
reduced product is formed after abstraction of the first bromine
atom and subsequent reduction, then this molecule gets a second
chance at cyclization when the second bromine reacts. In effect,
this group has one recycle built into it. In such double-barreled
applications, these aryl radicals have an advantage over more
traditional hydrogen abstracting radical precursors such as
preparative experiments should favor intramolecular hydrogen
transfer, but this effect is counterbalanced by the isotope effect
of changing from tin deuteride to tin hydride, which favors direct
reduction. Under our conditions, these two effects appear to
roughly cancel. Lower tin hydride concentrations might be
2
3
hypohalites, which are inherently “single-barreled” (that is,
each precursor only gets one shot at hydrogen abstraction).
o,o-Dibromo-p-methoxyphenyl ethers 17b and 31b of two
achieved either by slower syringe pumping or by changing the
bromide to an iodide,18 so it might be possible to increase the
yield of 29 closer to the maximum limit (73%) imposed by the
1
,6-hydrogen transfer.
The results of a another pair of secondary and tertiary
(
19) Keck, G. E.; Tafesh, A. M. Synlett 1990, 257.
(20) Radical additions to ketones: Zhang, W.; Dowd, P. Chem. ReV.
1993, 93, 2091.
substrates shown in entries 3a and 4a closely parallel the results
in entries 1a and 2. Reduction of 31a affords a 48% yield of
3
(
21) 5-exo-Cyclizations of alkoxy radicals are very rapid. Beckwith, A.
L. J.; Hay, B. P. J. Am. Chem. Soc. 1988, 110, 4415.
22) For examples of the “multiple-barreled” approach to cyclizations,
see: (a) Wilcox, C. S.; Gaudino, J. J. J. Am. Chem. Soc. 1986, 108, 3102.
b) Hirai, Y.; Hagiwara, A.; Terada, T.; Yamazaki, Y. Chem. Lett. 1987,
2417.
(23) Majetich, G.; Wheless, K. Tetrahedron 1985, 41, 7095.
2a (cyclized/reduced ratio 56/44), while the reduction of 33a
(
affords a 74% yield of 34a (cyclized/reduced ratio 80/20).
(
(18) Porter, N. A.; Magnin, D. R.; Wright, B. T. J. Am. Chem. Soc. 1986,
1
08, 2787.

3
146 J. Am. Chem. Soc., Vol. 118, No. 13, 1996
Curran and Xu
After the radical translocation reaction is complete, the PRT
group becomes a standard p-methoxyphenyl ether. These
groups are typically removed under mild oxidative conditions
with ceric ammonium nitrate. To demonstrate that removal was
possible, we treated three of our products under standard
1
0a
conditions with CAN (acetonitrile/water, 0 °C, 15 min). In
each case, a good yield of the expected primary alcohol was
isolated (eq 5).
Figure 4.
alcohol precursors of secondary radicals were prepared, and
these were reduced under the standard syringe pump conditions,
except that 2 equiv of tin hydride was used in place of 1 equiv.
In both cases (entries 1b and 3b), the yield of the cyclized
product increased significantly (for example, compare entries
1
7
a,b: 44% rises to 71%) as did the C/R ratio (53/47 rises to
9/21). These yields and ratios increased by approximately the
statistical amount that is expected if a radical has two equal
opportunities to be formed by 1,5-hydrogen transfer rather than
one. We also observed noticeable changes in the cis/trans ratios
of the cyclized products 29b and 32b. These changes might
originate because the translocated monobromoaryl radical
cyclizes with a different selectivity than the parent. A second
origin for this variation is that the intermediate monobromo
product is subject to a second radical translocation that can result
Conclusions
The results described in this paper show that it is possible to
invent PRT groups whose purpose is to generate radicals â to
the functional groups that they protect. The o-bromo-p-
methoxyphenyl ether PRT groups for generating radicals â to
alcohols have excellent protecting group characteristics. They
are easy to introduce and are removed under mild conditions.
They are superior protecting groups after translocation, being
among the most stable of any class of protecting groups for
in epimerization (see eq 4).5
e,6,8d
24
alcohols. Due to the aryl bromide, they have increased liability
as protecting groups prior to the radical translocation step
(especially toward reductive debromination or halogen-metal
exchange); however, even with the bromine present, these
groups will exhibit a good protection profile.
With respect to radical translocation reactions, the groups are
already useful, but further improvements are desirable. Prob-
lems arise due to relatively slow 1,5-hydrogen transfer, and due
to competing 1,6-hydrogen transfer. The problems with slow
On the basis of the rationale shown in Figure 4, we had hoped
that these dibromides would actually provide better than
statistical increases in the C/R ratio. Introduction of an
o-substituent larger than hydrogen might raise the energies of
unproductive conformers like 41 relative to 40, and in so doing
increase the efficiency of radical translocation relative to
bimolecular reduction. It is less obvious whether such a
substituent effect would alter the ratio of 1,5/1,6-hydrogen
transfer. If such an effect exists with the dibromides, it is not
very large, and it cannot be discerned from our data.
To better probe this idea, we prepared and reduced two
o-bromo-o′-methyl-p-methoxyphenyl ethers (31c and 33b). In
generation of a secondary radical, this group provides a
significantly higher yield of cyclized product and a higher C/R
ratio than the parent (entry 3c). Indeed, the isolated yield of
translocation/cyclization product 32c from this experiment is a
serviceable 76% and the C/R ratio is 83/17. This group also
proved superior for generating tertiary radicals (entry 4b), and
product 34b was isolated from 33b in 85% yield (C/R ratio
1
,5-hydrogen transfer can be minimized by lowering the tin
hydride concentration, but those with competing 1,6-hydrogen
transfer cannot. The parent o-bromo-p-methoxyphenyl group
shows low potential for generating simple secondary radicals
(
efficiencies ∼50%), but better potential for generating conju-
gated secondary and tertiary radicals (efficiencies ∼75%). The
o-substituted dibromo and o-bromo-o′-methyl analogs show
improved efficiencies for generating secondary radicals (g75%)
and quite good efficiencies for generating tertiary radicals
(g85%). It is not yet clear how these efficiencies in simple
acyclic systems will translate into more complex, conforma-
tionally restricted systems.
The original design premise of this group was our rate and
selectivity analogy between 5-exo-cyclization and 1,5-hydrogen
transfer shown in Figure 2. In retrospect, the results suggest
that this design premise was only partially correct. The
estimated rate constants of intramolecular hydrogen transfer
9
3/7). The o-methyl-substituted group is about as efficient as
the dibromide at hydrogen transfer despite its statistical
handicap. Larger o-substituents than methyl do not appear to
be advantageous, as reflected by the reductions of substrates
2
5
reactions of this class of radicals are at the lower end of the
2
6
useful scale for aryl radicals, and are considerably lower than
3
1d and 33c. The bulky 2-((tert-butyldimethylsilyl)oxy)-2-
the rate constant for 5-exo-cyclization of the analogous radical
propyl group gives about the same yields and C/R ratios as the
simple methyl group (entries 3d and 4c). This probably means
that these groups accelerate hydrogen transfer reactions enough
to significantly suppress direct reduction and that most of the
residual reduced product comes from the 1,6-hydrogen transfer
pathway. These results suggest that the o-methyl PRT group
should be superior to the parent for most applications. It is
about as good as the dibromide in efficiency, and it is
significantly simpler to use (only 1 equiv of tin hydride is
needed, and radical translocation after cyclization is not pos-
sible).
1
in Figure 2. However, the reduced rate of hydrogen transfer
relative to cyclization was anticipated on the basis of the earlier
(24) Greene and Wuts (ref 2) list this group as stable to the following
conditions: 3 N HCl, 100 °C; 3 N NaOH, 100 °C; H2, 1200 psi; O3, MeOH,
-
78 °C; LiAlH4; Jones reagent, and PCC.
(25) (a) Estimated rate constants (50 °C) for transfer of secondary
hydrogens are in the range of (1-5) × 106 s-1, and those of tertiary
7
-1
8
hydrogens are >10 s . (b) The estimates are arrived at by using 2 × 10
-
1
-1
m
s
as the rate constant for reaction of tributyltin deuteride with aryl
radicals. See ref 10b,d.
26) Aryl radical/solvent reactions are quite fast, so rates of 10 s or
better are needed for reactions to be preparatively useful.
6
-1
(

o-Bromo-p-methoxyphenyl Ethers
J. Am. Chem. Soc., Vol. 118, No. 13, 1996 3147
observations of Beckwith (see 3 f 4),10 and these relative rates
must be taken in context. Carbon-carbon π-bonds are much
weaker than carbon-hydrogen σ-bonds, so simple exothermicity
considerations suggest the guideline that cyclizations to alkenes
will usually be faster than analogous hydrogen transfer reactions.
Within the context of intramolecular hydrogen transfer reactions
in flexible substrates to aryl or vinyl radicals, we submit that
the rate constants for these radicals are quite high.
hydrogen transfer.5d One could then invent PRT groups that
reach even further down the chain to abstract hydrogens and
generate radicals.
The groups reported in this paper could become charter
members of a club of protecting groups that generate radicals
â to functional groups. Variations on the aryl group might allow
new options for introduction and cleavage or alter the stability
of the protecting group. Variations of the functional group
might allow for the generation of other classes of radicals, for
example, those â to amines or amides.
Though the rate analogy holds reasonably well, the selectivity
analogy does not. This analogy suggested that high 5-exo/6-
endo ratios in cyclizations would translate to analogously high
ratios of 1,5/1,6-hydrogen atom transfer. Substrate 1 cyclizes
exclusively by the 5-exo pathway, but its analog 3 gives only
about a 2/1 ratio of 1,5/1,6-hydrogen transfer. The selectivity
analogy is probably more tenous than the rate analogy because
the stereoelectronic effects that operate in cyclizations and
hydrogen transfers have different trends and origins. Radicals
Experimental Section
General Procedure for Radical Cyclization Reactions. Tributyltin
hydride (1.3-1.5 equiv containing 0.3 equiv of AIBN dissolved in
benzene) was added by syringe pump over a period of 10-15 h to a
solution of the substrate (1.0 equiv, 0.01 M) and AIBN (0.1 equiv) in
refluxing benzene. After evaporation of most of the solvent, wet ether
was added followed by addition of DBU (2.0 equiv) and a 1 M solution
2
7
prefer to attack alkenes with an approach angle of 109°, and
this angle is more easily attained in 5-exo-cyclizations than in
of I
2
in ether.17 A white precipitate formed, and stirring was continued
for 20 min. The mixture was filtered through a pad of silica gel. The
ether solution was concentrated, and the residue was purified by flash
chromatography.
General Procedures for Deuterium Labeling Experiments. The
substrate (1.0 equiv, 0.01 M), tributyltin deuteride (1.5 equiv), and
AIBN (0.2 equiv) were refluxed in benzene for 8 h. The crude mixture
6-endo-cyclizations. In contrast, intramolecular hydrogen trans-
fer reactions can only occur in an “endo” sense, and the favored
angle of 180° can be approached better in a 1,6-hydrogen
transfer than in a 1,5-hydrogen transfer.28 This analysis seems
to suggest that the ratio of rate constants for 5-exo/6-exo-
cyclization might be a better predictor for the 1,5/1,6-hydrogen
transfer rather than a 5-exo/6-endo cyclization ratio. However,
this is also not supported by the rates of cyclization 1/3 because
the 5-exo/6-exo (1/3) ratio rate constant10 is again considerably
higher than the 1,5/1,6-hydrogen transfer of our secondary
substrates.
was treated with DBU/I
2
as described in the general procedure. The
2
ether solution was concentrated, and the residue was subjected to H
NMR analysis without further purification to determine the sites of
deuteration.
Acknowledgment. We thank the National Institutes of
Health for funding this work.
The problem of competing 1,6-hydrogen transfer is not unique
to these PRT groups,5
d,6a
and this problem constitutes a
Supporting Information Available: Text describing the
experimental details and characterization of all compounds
reported in the paper (13 pages). This material is contained in
many libraries on microfiche, immediately follows this article
in the microfilm version of the journal, can be ordered from
the ACS, and can be downloaded from the Internet; see any
current masthead page for ordering information and Internet
access instructions.
potentially serious limitation that, depending on substituents,
may be difficult to overcome in some cases. Turning this
problem around, there is reasonable hope for using substituent
effects to increase the rate of 1,6-hydrogen transfer over 1,5-
(
27) (a) Beckwith, A. L. J.; Schiesser, C. H. Tetrahedron 1985, 41, 3925.
b) Spellmeyer, D. C.; Houk, K. N. J. Org. Chem. 1987, 52, 959.
28) This angle is much softer than the favored angle for radical
addition: (a) Dorigo, A. E.; Houk, K. N. J. Am. Chem. Soc. 1987, 109,
195. (b) Huang, X. L.; Dannenberg, J. J. J. Org. Chem. 1991, 56, 5421.
(
(
2
JA954150Z