Reactions of 1- and 2-Halo and 1,2-Dichloroadamantanes with Nucleophiles by the SRN1 Mechanism

Article abstract of DOI:10.1021/jo960965u

2-Bromoadamantane (2-BrAd) reacted in liquid ammonia under irradiation with diphenylphosphide (Ph₂P^-) ions whereas 2-chloroadamantane (2-ClAd) did not under the same experimental conditions. However, 2-ClAd yielded 2-(trimethylstannyl)adamantane in its photostimulated reaction with trimethylstannyl (Me₃Sn^-) ions. The compound 1-ClAd yielded the substitution product in a photostimulated slow reaction when the nucleophile is Ph₂P^- ion; the reaction occurs faster with the nucleophile Me₃Sn^- ion. All these reactions can be explained by the S_RN1 mechanism as they did not occur in the dark and were inhibited by p-dinitrobenzene when photostimulated. In competition experiments, 1-haloadamantane showed more reactivity than 2-haloadamantane. Either with Ph₂P^- or Me₃Sn^- ions, 1-BrAd is 1.4 times more reactive than 2-BrAd while 1-ClAd is 12 times more reactive than 2-ClAd with Me₃Sn^- ions. In the photostimulated reaction of 1,2-dichloroadamantane (7) with Ph₂P^- the monosubstitution products 1-adamantyldiphenylphosphine (64%) and 2-adamantyldiphenylphosphine (15%) were formed, isolated as the oxides. From these results, it appears that when 7 receives an electron, the 1-position fragments ca. four times faster than the 2-position. The disubstitution product was not formed with Ph₂P^- ions, but when 7 reacted with a nucleophile having less steric bulk such as a Me₃Sn^- ion, the 2-chloro-1-(trimethylstannyl)-adamantane and the disubstitution product 1,2-bis(trimethylstannyl)adamantane were formed. The formation of these products is explained in terms of the different rates of the electron transfer reactions of the radical anion intermediates.

Full text of DOI:10.1021/jo960965u

4406
J . Org. Chem. 1997, 62, 4406-4411
Rea ction s of 1- a n d 2-Ha lo- a n d 1,2-Dich lor oa d a m a n ta n es w ith
Nu cleop h iles by th e SRN1 Mech a n ism
Ana N. Santiago, Adriana E. Stahl, Gladis L. Rodriguez, and Roberto A. Rossi*
Departamento de Quı´mica Orga´nica, Facultad de Ciencias Quı´micas, Universidad Nacional de Co´rdoba,
A.P. 4 - C.C. 61 - 5000 Co´rdoba, Argentina
Received May 24, 1996 (Revised Manuscript Received J anuary 16, 1997X)
2-Bromoadamantane (2-BrAd) reacted in liquid ammonia under irradiation with diphenylphosphide
(Ph2P-) ions whereas 2-chloroadamantane (2-ClAd) did not under the same experimental conditions.
However, 2-ClAd yielded 2-(trimethylstannyl)adamantane in its photostimulated reaction with
trimethylstannyl (Me3Sn-) ions. The compound 1-ClAd yielded the substitution product in a
photostimulated slow reaction when the nucleophile is Ph2P- ion; the reaction occurs faster with
the nucleophile Me3Sn- ion. All these reactions can be explained by the SRN1 mechanism as they
did not occur in the dark and were inhibited by p-dinitrobenzene when photostimulated. In
competition experiments, 1-haloadamantane showed more reactivity than 2-haloadamantane.
Either with Ph2P- or Me3Sn- ions, 1-BrAd is 1.4 times more reactive than 2-BrAd while 1-ClAd is
12 times more reactive than 2-ClAd with Me3Sn- ions. In the photostimulated reaction of 1,2-
dichloroadamantane (7) with Ph2P- the monosubstitution products 1-adamantyldiphenylphosphine
(64%) and 2-adamantyldiphenylphosphine (15%) were formed, isolated as the oxides. From these
results, it appears that when 7 receives an electron, the 1-position fragments ca. four times faster
than the 2-position. The disubstitution product was not formed with Ph2P- ions, but when 7 reacted
with a nucleophile having less steric bulk such as a Me3Sn- ion, the 2-chloro-1-(trimethylstannyl)-
adamantane and the disubstitution product 1,2-bis(trimethylstannyl)adamantane were formed. The
formation of these products is explained in terms of the different rates of the electron transfer
reactions of the radical anion intermediates.
Several alkyl halides have been found to react with
process has also been proposed in the reaction of 1-io-
doadamantane (1-IAd) with carbanions,2,4 PhS-,5 PhSe-,5
and PhTe- 5 ions. Recently, it has been demonstrated
that the photostimulated reaction of 1-IAd with arenethi-
olate ions occurs by a nonchain SRN1 reaction.6 On the
other hand the trimethylstannylation of 1-BrAd and
1-IAd with NaMe3Sn in THF at 0 °C proceeds predomi-
nantly by a free radical process.7 Other alkyl halides8
and other bridgehead halides9 have been proposed to
react with Me3Sn- ions by electron transfer.
nucleophiles by the radical nucleophilic substitution, or
S
RN1 mechanism.1 The main steps are shown in eqs 1-3.
The SRN1 reactions of dihalo bridgehead compounds
with nucleophiles give either the monosubstitution or
disubstitution product depending on the radical type, the
nature of the nucleofugal group, or the nucleophile.10,11
The monosubstitution product is not an intermediate in
the formation of the disubstitution one. For example, in
the SRN1 photostimulated reaction of 1,3-dihaloadaman-
When there is no spontaneous electron transfer (ET)
from the Nu- to the substrate RX, the initiation step can
occur under photostimulation1 or with FeBr2 in DMSO.2
The alkyl radical R• thus formed couples with the Nu-,
yielding a new radical anion (RNu)-• (eq 2), which by an
intermolecular dissociative ET3 to the substrate gives the
substitution product and the alkyl radical R• to continue
the chain propagation cycle (eq 3).
(4) Rossi, R. A.; Pierini, A. B.; Borosky, G. L. J . Chem. Soc., Perkin
Trans. 2 1994, 2577.
(5) Palacios, S. M.; Alonso, R. A.; Rossi, R. A. Tetrahedron 1985,
41, 4147.
The alkyl halides that react by the SRN1 mechanism
are mainly those that have a relatively low reactivity
toward polar nucleophilic substitution reactions. For
instance, 1-haloadamantanes as well as other bridgehead
halides react with Ph2P- ions by this mechanism.1 The
(6) Ahbala, M.; Hapiot, P. K.; Houmam, A.; J ouini, M.; Pinson, J .;
Save´ant, J . M. J . Am. Chem. Soc. 1995, 117, 11488.
(7) (a) Smith, G. F.; Kuivila, H. G.; Simon, R.; Sultan, L. J . Am
Chem. Soc. 1981, 103, 833. (b) Duddeck, H.; Islam, M. R. Chem. Ber.
1984, 117, 565.
(8) (a) Ashby, E. C.; DePriest, R. N.; Su, W. Y. Organometallics 1984,
3, 1718. (b) Ashby, E. C.; Su, W. Y.; Pham, T. N. Organometallics 1985,
4, 1493. (c) San Filippo, J .; Silbermann, J .; Fagan, P. J . J . Am. Chem.
Soc. 1978, 100, 4834. (d) Lee, K-W.; San Filippo J . Organometallics
1983, 2, 906.
(9) (a) Ashby, E. C.; DePriest, R. N.; Su, W. Y. Organometallics 1984,
3, 1718. (b) Alnajjar, M. S.; Kuivila, H. G. J . Org.Chem. 1981, 46, 1053.
(c) Ashby, E. C.; Sun, X.; Duff, J . L. J . Org.Chem. 1994, 59, 1270. (d)
Adcock, W.; Clark, C. I. J . Org. Chem. 1995, 60, 723.
(10) Palacios, S. M.; Santiago, A. N.; Rossi, R. A. J . Org. Chem. 1984,
49, 4609.
X Abstract published in Advance ACS Abstracts, April 1, 1997.
(1) For reviews see (a) Rossi, R. A.; Pierini A. B.; Palacios, S. M.
Adv. Free Radical Chem. (Greenwich, Conn.) 1990, 1, 193. Rossi, R.
A.; Pierini, A. B.; Palacios, S. M. J . Chem. Ed. 1989, 66, 720. (b) Rossi,
R. A.; Pierini A. B.; Pen˜e´n˜ory, A. B. The Chemistry of Functional Group,
Patai, S., Rappoport, Z., Eds.,Wiley: Chichester, 1995; Supl. D2, Ch
24, p 1395.
(2) Nazareno, M. A.; Rossi, R. A. J . Org. Chem. 1996, 61, 1645.
(3) Save´ant, J . M. Adv. Electron Transfer Chem. 1994, 4, 53, and
references cited therein.
(11) Santiago, A. N.; Iyer, V. S.; Adcock, W.; Rossi, R. A. J . Org.
Chem. 1988, 53, 3016.
S0022-3263(96)00965-6 CCC: $14.00 © 1997 American Chemical Society

SRN1 Reaction of Haloadamantanes with Nucleophiles
J . Org. Chem., Vol. 62, No. 13, 1997 4407
tanes with Ph2P- ions mainly the disubstitution product
is formed.12 In the reaction of 1,3-dihaloadamantanes
with NaMe3Sn in THF at 0 °C, the products depend on
the leaving groups. Thus, when the halides are iodides
or bromides the reaction proceeds predominantly with
the formation of 1,3-adamantene, but with chlorides or
1-chloro-3-bromo, the disubstitution product is formed
through the SRN1 mechanism.13 Besides, Adcock et al.
reported that trimethylstannylation of dihalo bridgehead
compounds in THF at 0 °C proceed between two path-
ways (radical or polar process).14
Ta ble 1. Rea ction of 1- a n d 2-Ha loa d a m a n ta n es w ith
P h 2P - a n d Me3Sn - Ion s in Liqu id Am m on ia
products, % yielda
conditions X- subst product
substrate
nucleophile
expt
(×103 M)
(×103)
1b 1-BrAd (7.20) Ph2P- (5.80) hν, 20 min 100 1-AdP(O)Ph2,
70
2
1-ClAd (2.67) Ph2P- (2.95) hν, 4 h
39 1-AdP(O)Ph2,
29
3
4
5
6
7
8
9
2-ClAd (2.67) Ph2P- (3.40) hν, 4 h
2-ClAd (2.67) Ph2P- (3.40) dark, 4 h
2-BrAd (3.30) Ph2P- (3.40) dark, 1 h
17 1-AdP(O)Ph2
c
<3d
-
e
23 2-AdP(O)Ph2
2-BrAd (3.30) Ph2P- (3.30) dark, 1 hf <1d
-
2-BrAd (3.31) Ph2P- (3.33) hν, 1 h
2-BrAd (3.30) Ph2P- (3.30) hν, 1 hf
1-ClAd (1.63) Me3Sn- (1.66) hν, 20 min
80 2-AdP(O)Ph2, 64
10 2-AdP(O)Ph2, 4
Although the 2-position of the adamantane system is
a secondary carbon atom of a cyclohexane system, it
shows a lower reactivity than the 1-position toward
nucleophiles in polar reactions. It has been reported that
1-adamantyl tosylate solvolyzes about 105 times faster
than 2-adamantyl tosylate.15 Besides, chloro and bromo
cyclohexanes and derivatives react under irradiation with
Ph2P- ions in liquid ammonia by the SRN1 mechanism.16
e
e
1-AdSnMe3, 93
10 1-ClAd (1.67) Me3Sn- (1.60) dark,
1-AdSnMe3
c
20 min
11 1-ClAd (1.67) Me3Sn- (1.70) hν, 20 mind
12 2-ClAd (1.66) Me3Sn- (1.66) hν, 75 min
e
e
e
1-AdSnMe3
c
2-AdSnMe3, 94
13 2-ClAd (1.68) Me3Sn- (1.69) dark,
75 min
2-AdSnMe3
c
a
Yields are based on substrate concentration and determined
by GLC using the internal standard method. See ref 29. c None
b
The reaction of 2-chloro-5-iodoadamantane (E and Z)
with LiMe3Sn in THF at 0 °C gave mainly the reduction
product 2-ClAd and the monosubstitution compound
2-chloro-5-(trimethylstannyl)adamantane, but in the re-
action with 2-bromo-5-iodoadamantane (E and Z), the
disubstitution product is obtained in high yield. 2-Bromo-
5-fluoroadamantane (E and Z) gave the monosubstitution
product 2-(trimethylstannyl)-5-fluoroadamantane. On
the basis of trapping experiments with dicyclohexylphos-
detected. The substrate was recovered unchanged. e Not quanti-
d
fied. f p-DNB was added (20 mol %).
slow dark reaction was completely inhibited by p-DNB,
indicating that there is a spontaneous ET from the
nucleophile to the substrate.
phine, these reactions were proposed to occur by the SRN
1
mechanism.17
Tabushi et al. reported, in the free-radical halogenation
of adamantane, that the bridgehead position has a higher
reactivity than the bridge position (or secondary position).
They also claim that the 1-adamantyl radical is formed
easily and has a longer lifetime than the 2-adamantyl
radical, suggesting that it may be considerably more
stable than the latter radical.18
On the other hand, in the reaction of 2-ClAd with
Ph2P- ions only 17% yield of chloride ions was deter-
mined after 4 h of irradiation without formation of 2. No
dark reaction was observed (Table 1, experiments 3 and
4). 1-ClAd (3) reacts with Ph2P- to give 29% yield of the
substitution product (Table 1, experiment 2). Thus
1-ClAd results to be more reactive than 2-ClAd in
separated experiments.
On account of the previous findings we desired to
undertake the study of the reactions of 1-halo-, 2-halo-,
and 1,2-dihaloadamantanes with Ph2P- and Me3Sn- ions
under SRN1 conditions in order to get more insights on
the relative SRN1 reactivities of the 1- and 2-positions of
halo- and dihaloadamantanes.
Under irradiation (20 min), 3 reacts with Me3Sn- ions
to give 93% yield of the substitution product 4 (eq 5). No
reaction was found in the dark, and the photostimulated
reaction was inhibited by p-DNB (Table 1, experiments
9-11).
Resu lts a n d Discu ssion
Reaction s of 1- or 2-Haloadam an tan es with Diph e-
n ylp h osp h id e a n d w ith Tr im eth yltin Ion s. 2-BrAd
(1) reacts slowly with Ph2P- ions in liquid ammonia
under dark conditions to afford the substitution product
2, isolated as the oxide. The reaction is accelerated by
irradiation (eq 4). The facts that the reactions are light
catalyzed and that the photostimulated and dark reac-
tions are inhibited by p-dinitrobenzene (p-DNB), a well
known inhibitor of SRN1 reactions,1 suggest that they
occur by this mechanism (Table 1, experiments 5-8). The
In contrast, 2-ClAd (5) requires a longer irradiation
time (75 min) to yield 94% of substitution product 6 with
Me3Sn- ions. In this system there is no dark reaction
(Table 1, experiments 12 and 13) (eq 6).
(12) Lukach, A. E.; Santiago, A. N.; Rossi, R. A. J . Phys. Org. Chem.
1994, 7, 610.
(13) Adcock, W.; Clark, C. I J . Org. Chem. 1993, 58, 7341.
(14) Adcock, W.; Iyer, V. S. J . Org. Chem. 1985, 50, 3706.
(15) Bentley, W.; Carter, G.; Robert, K. J . Org. Chem. 1984, 49, 5183.
(16) Nazareno, M. A., Palacios, S. M.; Rossi, R. A. J . Phys. Org.
Chem. 1993, 6, 421.
(17) Adcock, W.; Clark, C. I.; Trout, N. A. Tetrahedron Lett. 1994,
35, 297.
(18) Tabushi, I.; Aoyama, Y.; Kojo, S.; Hamuro, J .; Yoshida, Z. J .
Am Chem. Soc. 1972, 94, 1177.
The fact that the reactions of 3 and 5 with Me3Sn- ions
are light catalyzed is in agreement with the expected
RN1 mechanism. The different irradiation times neces-
sary to complete the reaction with this nucleophile (20
min for 3 and 75 min for 5) show that 3 is more reactive
than 5 toward Me3Sn- ions under these experimental
conditions. The behavior observed with Ph2P- ions,
which do not react with 5 under irradiation but react with
S

4408 J . Org. Chem., Vol. 62, No. 13, 1997
Santiago et al.
Ta ble 2. Com p etition Exp er im en ts of 1- a n d 2-Ha loa d a m a n ta n es w ith P h 2P - a n d Me3Sn - Ion s in Liqu id Am m on ia
% products, yield
expt
1-XAd (×103 M)
X ) Br (3.33)
X ) Br (1.75)
X ) Br (1.58)
2-XAd (×103 M)
X ) Br (3.33)
X ) Br (1.71)
X ) Br (1.60)
nucleophile (×103 M)
1-AdNu
2-AdNu
k1-XAd/k2-XAd
1.4
1.5
1.3
1.4 ( 0.1 (Average)
13.8
10.2
12.0
1a
2c
3c
Ph2P- (3.33)
Me3Sn- (1.70)
Me3Sn- (1.69)
27b
58
51
19b
45
44
4a
5a
6a
X ) Cl (1.70)
X ) Cl (1.70)
X ) Cl (1.68)
X ) Cl (1.67)
X ) Cl (5.00)
X ) Cl (8.33)
Me3Sn- (1.83)
Me3Sn- (1.80)
Me3Sn- (1.83)
80
59
61
11
8.4
7.6
12 ( 1.2 (Average)
a
b
Irradiation time 75 min. Quantified as the oxides. c Irradiation time 15 min.
Ta ble 3. Rea ction s of 7 w ith P h 2P - a n d Me3Sn - Ion s in
Liqu id Am m on ia
2-BrAd, reveals that the initiation step of the mechanism
is not efficient between Ph2P- and 5.
7
nucleophile
products,
% yielda
Com p etition Rea ction s. In order to determine the
relative reactivity of 1-haloadamantanes vs 2-haloada-
mantanes in SRN1 reactions, we studied their reactivity
toward Me3Sn- and Ph2P- ions in competition experi-
ments. All the reactions were performed with excess of
the substrates. The relative reactivity was estimated as
in previous work.19
When a solution containing substrates 1- and 2-BrAd
is allowed to compete for Ph2P- ions under irradiation,
from the amount of products obtained it is possible to
determine that 1-BrAd is 1.4 times more reactive than
2-BrAd (Table 2, experiment 1). Similar relative reac-
tivities were found in competitive photostimulated reac-
tions of 1-BrAd and 2-BrAd toward Me3Sn- ions (Table
2, experiments 2 and 3).
expt
(×103 M)
(×103 M)
conditions
1b
2
3
4
5
6
7
8f
2.33
2.33
2.33
1.70
1.71
1.65
1.69
1.20
Ph2P- (4.66)
Ph2P- (4.66)
Ph2P- (4.66)
Me3Sn- (3.66)
Me3Sn- (3.33)
Me3Sn- (3.40)
Me3Sn- (11)
Me3Sn- (1.25)
hν, 4 h
2, 15; 8, 64
2,c 8c
2, 13; 8, 47
17, 54; 18, 26
17,c 18c
17,c 18c
dark, 4 h
hν, 4 hd,e
hν, 2 h
dark, 2 h
hν, 2 hd
hν, 4 h
17, 50; 18, 30
17, 93; 18, <4
hν, 2 h
a
Yields are based on substrate concentration and determined
b
by GLC using the internal standard method. Duplicate experi-
ments. Small amounts (<4%) of adamantane, 1-ClAd and 2-ClAd
d
were formed. c None detected. p-DNB was added (20 mol %).
e 20% of 7 was recovered. f The substrate was 17.
On the other hand, when 7 reacts with Ph2P- ions in
liquid ammonia, 92% yield of chloride ion release is
determined after 240 min irradiation (based on two
chlorines per molecule), together with the monosubsti-
tuted products 2 and 8 quantified as the oxides in 15%
and 64% yields, respectively (eq 7). No disubstitution
product 1,2-bis(diphenylphosphinyl)adamantane was
found. There was no reaction in the dark after the same
reaction time, and the photostimulated reaction was
slightly inhibited by p-DNB under this experimental
conditions (Table 3, experiments 1-3).
A relative reactivity of 12 was determined in the
competitive photostimulated reactions of 1-ClAd and
2-ClAd toward Me3Sn- ions (Table 2, experiments 4-6).
Either with nucleophiles Me3Sn- or Ph2P-, 1-haloada-
mantane is more reactive than 2-haloadamantane in
S
RN1 reactions, the difference in reactivity being higher
with the poorest leaving group.
The relative reactivity of two substrates toward a given
nucleophile will depend on the relative electron acceptor
capability of the substrates toward the electron transfer
reaction of eq 3, given a propagation cycle in which there
are no other competing reactions. The observed relative
reactivity of the compounds 1-haloadamantane and 2-ha-
loadamantane with the same nucleophile (Ph2P- or
Me3Sn-) is thus related to their reduction potential and
is in agreement with the calculated σ* MO energy of the
C-X bond of the substrates.20
Reaction s of 1,2-Dich lor oadam an tan e with Diph e-
n ylp h osp h id e a n d w ith Tr im eth yltin Ion s. It has
been reported that adamantene is the intermediate in
the reaction of LiPh2P in THF with 1,2-diiodo- and 1,2-
dibromoadamantane but not with 1,2-dichloroadaman-
tane (7), although 1- and 2-adamantyldiphenylphosphine
are formed in all cases, isolated as the oxides.21
These results suggest that the reaction occurs by the
RN1 mechanism. The nucleophile transfers one electron
S
to the substrate in a light-catalyzed initiation step to give
the haloadamantyl radicals (9 and 10) (eq 8) which then
couple with the nucleophile to form the radical anion
intermediates (11-• and 12-•) (eq 9). These radical anions
fragment by dissociative intramolecular ET to the C-Cl
bond to yield radicals 13 and 14 (eq 10). These radicals
do not couple with the nucleophile probably due to a
steric effect, but are reduced to products 15 or 16 (eq
11),22 which after oxidation give the products 2 and 8.
(19) The equation used in the relative reactivity determination of
pairs of substrates vs a nucleophile is:
ks1 ln [S1]0/[S1]t
)
ks2
ln [S2]0/[S2]t
where [S1]0 and [S2]0 are initial concentrations, and [S1]t and [S2]t are
concentrations at time t of both substrates. See Bunnett, J . F. Investigation
of Rates and Mechanisms of Reactions, 3rd ed.; Lewis, E. S., Ed.; Wiley-
Interscience: New York, 1974; Part 1, p 159.
(20) The LUMO value of the C-Br σ* MO in 1-BrAd (0.767 eV) is
lower than the C-Br σ* MO of 2-BrAd (0.867 eV). 1-ClAd (1.446 eV)
has a C-Cl σ* MO energy lower than the C-Cl σ* MO of the 2-ClAd
(1.548 eV). The calculations were carried out with the AM1 method.
(21) Gillespie, D. G.; Walker, B. J . J . Chem. Soc., Perkin Trans. 1
1983, 1689.
(22) A possibility for the formation of the reduction products is the
ET from the nucleophile to the radical followed by protonation.
However, although liquid ammonia is not a good hydrogen donor,
radicals reacting poorly toward nucleophiles can be reduced by the
solvent. As we do not have experimental evidence to favor one or the
other mechanism we cannot state which of them occurs in our system.

SRN1 Reaction of Haloadamantanes with Nucleophiles
J . Org. Chem., Vol. 62, No. 13, 1997 4409
that 17 is not an intermediate in the formation of 18.
Likewise, 17 fails to react with Me3Sn- ions in liquid
ammonia (2 h irradiation, 18 < 4%) (Table 3, experiments
7 and 8).
The fact that 17 does not react under irradiation with
Me3Sn- ions but 2-ClAd does could be due to the well
known powerful σ-electron donor capability of the Me3-
Sn group that would increase the LUMO energy of the
C-Cl bond.
These results can be explained in terms of the SRN1
mechanism. Under irradiation, 7 gives the radicals 9 and
10 in the sense of eq 8. The radicals 9 and 10 couple
with the nucleophile to give the radical anions intermedi-
ates 17-• and 19-• (eq 14).
These radical anions intermediates show different
behavior. The radical anion 19-• gives only the intramo-
lecular ET to form radical 20 and chloride ion. Radical
20 couples with the nucleophile to give ultimately the
disubstitution product 18 (eq 15). The monosubstitution
product with retention of chlorine formed by intermo-
lecular ET is not observed. In contrast, the radical anion
intermediate 17-• has two competitive reactions, in-
tramolecular ET to give radical 21, which ultimately
gives the disubstitution product 18 (eq 16), or intermo-
lecular ET to give 17, which is not an intermediate to
form 18 (eq 17).
The fact that no monosubstitution products 11 or 12
with retention of chlorine were found may indicate that
the intermolecular ET to the substrate (eq 12) does not
compete with the intramolecular ET to give radicals 13
and 14 (eq 10).
Taking into account that the yield of 16 is four times
higher than that of 15 it seems that when 7 receives an
electron it fragments four times faster at the 1-position
to give radical 10 than at the 2-position to give radical 9
(eq 8). On the other hand, 2-ClAd does not react with
Ph2P- ions under irradiation, probably due to an inef-
ficient photostimulated initiation step. However, the
radical anion intermediate 12-• has a fast intramolecular
ET rendering finally the product observed.
In order to see if the lack of formation of the disubsti-
tution product is due to the steric bulk of the radicals 13
and 14, and that of the incoming nucleophile, we studied
the reaction of 7 with a nucleophile with less steric bulk
than Ph2P- ions. Estimation of the steric bulk of the
trimethyltin group indicates that it is very similar to that
of the chlorine and smaller than bromine groups.23
When 7 reacts under photostimulation with Me3Sn-
ions mainly the monosubstitution product with retention
of chlorine, (2-chloro-1-adamantyl)trimethyltin (17) (54%),
and the disubstitution product, 1,2-bis(trimethystannyl)
adamantane (18) (26%), were formed (eq 13). The
reaction does not occur in the dark, and the photostimu-
lated reaction is inhibited by p-DNB (Table 3, experi-
ments 4-6).
If the rate of formation of radicals 9 and 10 by
photostimulated intermolecular ET from the nucleophiles
(Ph2P- or Me3Sn- ions) or by any radical anion interme-
diate is the same (9:10 ≈ 15:64 determined by product
analysis of the reaction with Ph2P- ions), the yield of 18
formed through eqs 14 and 15 should be ≈15%, and ≈11-
15% formed through eqs 14 and 16 in order to account
for the ≈26-30% yield of 18.
As the substitution product 17 did not give 18, the ratio
of the competing reactions of radical anion 17-., the
intramolecular ET to give 18 (eq 16, ≈11-15%) and the
intermolecular ET to give 17 (eq 17, ≈50-54%), is
approximately 4, on the basis of product analysis.
Con clu sion s
These results show that 2-ClAd does not react under
irradiation with Ph2P- ions, but Me3Sn- does, giving good
yields of the substitution product, which indicates that
Me3Sn- ions are more reactive in the initiation step of
the SRN1 mechanism. A more reactive substrate, such
as 2-BrAd, reacts with Ph2P- ions under irradiation.
The same relation of products 17 and 18 was found
when the reaction was performed with a large excess of
Me3Sn- ions and longer irradiation times (4 h), indicating
(23) Davis, D. D.; Surmatis, A. J .; Robertson, G. L. J . Organomet.
Chem. 1972, 46, C9.

4410 J . Org. Chem., Vol. 62, No. 13, 1997
Santiago et al.
By competition experiments with Me3Sn- ions, 1-ClAd
is 12 times more reactive than 2-ClAd, and by competi-
tion experiments with Ph2P- ions or Me3Sn- ions, 1-BrAd
is 1.4 times more reactive than 2-BrAd, indicating that
the 1-position is more reactive than the 2-position, and
this difference is higher with the poorest leaving group.
When 7 receives an electron, two positions are reactive,
but the fragmentation rate at the 1-position is ca. 4 times
faster than at the 2-position, as suggested in the reactions
with Ph2P- ions. The radicals formed by intramolecular
ET do not couple with the nucleophile probably due to
the steric bulk of both the nucleophile Ph2P- and the
diphenylphosphinyl moiety of the radical intermediates,
thus only the monosubstitution products are formed.
In the reaction of 7 with Me3Sn- ions, a less sterically
bulky nucleophile, the disubstitution product is obtained,
together with the monosubstitution product with reten-
tion of chlorine 17. It is interesting to note that similar
radical anions have different rates in ET reactions; thus
metal under nitrogen. Ph3P (1 mmol) and Na metal (2 mmol)
were added to form Ph2P- ions, and t-BuOH (1 mmol) was
added to neutralize the amide ions formed. To this solution
was added 1 mmol of substrate and then irradiated for 1 h.
The reaction was quenched by adding NH4NO3 in excess, and
the ammonia was allowed to evaporate. The residue was
dissolved with water and then extracted with diethyl ether
and with methylene chloride, too. The products were oxidized
with H2O2 and then quantified by GLC using the internal
standard method. In another experiment the product was
oxidized with H2O2, and 2 was isolated as a white solid after
chromatography on silica gel, eluted with diethyl ether, and
recrystallized from benzene: mp 209-210 °C (lit21 mp 209-
211 °C).
P h otostim u la ted Rea ction of 1,2-Dich lor oa d a m a n ta n e
w ith P h 2P - Ion s in Liqu id Am m on ia . The procedure was
similar to that for the previous reaction, except that 1,2-
dichloroadamantane was used as substrate. In this case the
reaction was irradiated for 4 h. The products were oxidized
and then quantified by GLC with the internal standard
method. The products 221 and 829 were compared with
authentic samples.
P h otostim u la ted Rea ction of 1-ClAd w ith Me3Sn - Ion s
in Liqu id Am m on ia . The following procedure is representa-
tive of these reactions. The equipment used has been de-
scribed for the reaction with Ph2P- ions. Nucleophile Me3Sn-
was prepared in liquid ammonia (lemon yellow solution) from
Me3SnCl (0.5 mmol) and Na metal (1.2 mmol, 20% excess).
The 1-ClAd (0.5 mmol dissolved in 1 mL of anhydrous diethyl
ether) was added, and then irradiated for 20 min. The
procedure was as usual. After workup, 1-(trimethylstannyl)-
adamantane14,30,31 was purified by sublimation (liquid air
cooling, 52-3 °C) to afford white crystals: mp 56-7 °C (lit.14,30
mp 56-7 °C). 1H NMR (CDCl3) δ -0.05 (s, 9 H, SnMe3; J Sn-H
47.57 and 49.81 Hz); 1.75-2.20 (m, 15 H).
the intramolecular ET of the radical anions 11-•, 12-•
,
and 19-• to the C-Cl bond is much faster than the
intermolecular ET, and none of the monosubstitution
products with retention of chlorine are obtained. How-
ever, in the radical anion 17-•, the intermolecular ET is
ca. 4 times faster than the intramolecular ET under our
experimental conditions.24
P h otostim u la ted Rea ction of 2-ClAd w ith Me3Sn - Ion s
in Liqu id Am m on ia . Under the conditions described above,
2-(trimethylstannyl)adamantane was formed after 75 min of
irradiation and isolated as white crystals by column chroma-
tography with light petroleum ether on silica gel: mp 37-8
°C (lit.30,31 mp 37-8 °C). 1H NMR (CDCl3) δ 0.05 (s, 9 H,
SnMe3; J Sn-H 47.60 and 50.00 Hz); 1.65-2.08 (m, 15 H).
Exp er im en ta l Section
Gen er a l Meth od s. Irradiation was conducted in a reactor
equipped with two 250-W UV lamps emitting maximally at
350 nm (Philips Model HPT, water-refrigerated). Column
chromatography was performed on silica gel (70-270 mesh
ASTM).
P h otostim u la ted Rea ction of 1,2-Dich lor oa d a m a n ta n e
w ith Me3Sn - Ion s in Liqu id Am m on ia . Using the proce-
dure described above, 1,2-dichloroadamantane (0.5 mmol) was
allowed to react with Me3Sn- ions (1 mmol). The residue was
column chromatographed on silica gel and eluted with light
petroleum ether. 1,2-Bis(tr im eth ylsta n n yl)a d a m a n ta n e:
It was isolated as a white solid and carefully sublimed (38-
42 °C) and recrystallized from acetone keeping it in a freezer:
mp 52-3 °C. 1H NMR (CDCl3) δ 0.00 (s, 9 H, SnMe3) and 0.10
(s, 9 H), J Sn-H not obsd; 1.68-2.27 (m, 14 H). 13C NMR (CDCl3)
δ -11.06 (J C-Sn 271.88 and 284.72 Hz), -7.30 (J C-Sn 278.32
and 291.29 Hz), 29.18 (J C-Sn 50.30 Hz), 29.25 (J C-Sn 47.91 Hz),
29.71 (J C-Sn not obsd), 33.31 (J C-Sn 55.36 Hz), 36.84 (12.82 Hz),
38.25 (J C-Sn 6.58 Hz), 40.45 (J C-Sn 8.69 Hz), 40.63, 45.66 (J C-Sn
69.54 Hz), 45.74 (J C-Sn 400.27 and 418.87 Hz). Mass spectra
similar to report.27 Anal. Calcd for C16H32Sn2: C 41.61, H
6.98. Found: C 41.52, H 7.07. 2-Ch lor o-1-(tr im eth ylsta n -
n yl)a d a m a n ta n e: a solid purified by sublimation: mp 65.5-
66.5 °C. 1H NMR (CDCl3) δ 0.05 (s, 9 H, J Sn-H 47.6 and 51.4
Hz); 1.60-2.51 (m, 13 H); 4.60 (s, 1 H). 13C NMR (CDCl3) δ
-11.08 (J C-Sn 297.80 and 313.00 Hz), 27.38 (J C-Sn 42.15 Hz),
28.30 (J C-Sn 42.89 Hz), 31.15, 35.22 (J C-Sn 7.40 Hz), 36.36,
36.73 (J C-Sn 29.94 Hz), 37.85 (J C-Sn 6.54 Hz), 38.44, 42.74
(J C-Sn 13.18 Hz), 75.39 ppm (J C-Sn 23.65 Hz). MS: m/z (rel
intensity) M+ not obsd 41 (7.6), 57 (21.6), 79 (23.4), 91 (31.02),
92 (100), 93 (28.4), 105 (18.8), 119 (18.1), 134 (82.0), 165 (3.74)
Ma ter ia ls. Reagents 2-ClAd (98%, J ohnson Matthey-Alfa
Products), 1-BrAd, 1-ClAd, and 2-BrAd (98%, Aldrich), 1-hy-
droxyadamantane (Sigma), and p-DNB (Fluka) were used as
received. Ph2P- ions were prepared from Ph3P (Fluka) and
Na metal in liquid ammonia. Me3Sn- ions were prepared from
Me3SnCl (Fluka) and Na metal in liquid ammonia.25
Syn t h esis of R ea ct a n t s. 4-Protoadamantanone was
prepared from 1-hydroxyadamantane as described26 by rear-
rangement of bridgehead alcohol and then purified by chro-
matography on aluminum oxide (6% w/w) and eluted with
petroleum ether: mp 201-2 °C (lit26 mp 202-204 °C). The
1
IR26 and H NMR27 are identical to those reported. Synthesis
of 1,2-dichloroadamantane: The reaction between 4-protoada-
mantanone and HCl/ZnCl2 was carried out as described.28 The
1,2-dichloroadamantane was isolated as a white solid after
chromatography on silica gel and eluted with petroleum
ether: mp 186-188 °C (lit28 mp 186-7 °C).
P h otostim u la ted Rea ction of 2-Br Ad w ith P h 2P - Ion s
in Liqu id Am m on ia . The following procedure is representa-
tive of all the reactions. Into a three-necked, 500-mL, round-
bottomed flask equipped with a cold finger condenser charged
with dry ice-ethanol, a nitrogen inlet, and a magnetic stirrer
were condensed 300 mL of ammonia previously dried with Na
(24) We tried to theoretically explain the difference in behavior
between radical anions 17-• and 19-•. However, PM3 calculations do
not show an energy difference between the SOMO and the C-Cl σ*
MOs of both intermediates.
(25) Yammal, C. C.; Podesta´, J . C.; Rossi, R. A. J . Organomet. Chem.
1996, 509, 1.
(26) Majerski, Z.; Hamersak, Z. Org. Synth. 1979, 59, 147.
(27) Momose, T.; Itooka, T.; Muraoka, O. Synth. Commun. 1984,
14, 147.
(29) Palacios, S. M.; Santiago, A. N.; Rossi, R. A. J . Org. Chem. 1982,
47, 4654.
(30) Kuivila, H. G.; Cosidine, J . L.; Sarma, R. H.; Mynott, R. J . J .
Organomet. Chem. 1976, 111, 179.
(31) Doddrell, D.; Burfitt, I.; Kitching, W.; Bullpitt, M.; Lee, C. H.;
Mynott, R. J .; Cosidine, J . L.; Kuivila, H. G.; Sarma, R. H. J . Am. Chem.
Soc. 1974, 96, 1640.
(28) Abdel-Sayed, A. N.; Bauer, L. Tetrahedron 1988, 44, 1873.

SRN1 Reaction of Haloadamantanes with Nucleophiles
J . Org. Chem., Vol. 62, No. 13, 1997 4411
135 (14.0), 185 (13.0). Anal. Calcd for C13H23ClSn: C 46.82,
H 6.95, Cl 10.63. Found: C 46.83, H 7.00, Cl 10.87.
Reaction of Differ en t Su bstr ates with P h 2P - or Me3Sn -
Ion s in th e Da r k . The described procedure was followed,
except that the reaction flask was wrapped with aluminum
foil.
excess. The reaction mixture was extracted with diethyl ether
and water. Substitution products were quantified by GLC
with the internal method. Relative reactivity of 1-BrAd vs
2-BrAd: The nucleophiles (Ph2P- or Me3Sn-) were prepared
in liquid ammonia, and the competition reactions were as
carried out as before.
In h ibited Rea ction of Differ en t Su bstr a tes w ith P h 2P -
or Me3Sn - Ion s. The procedure was similar to that for the
previous reaction, except that 20 mol % of p-DNB were added
to the solution of nucleophile prior to substrate addition.
Com p etition Exp er im en ts. Relative reactivity of 1-ClAd
vs 2-ClAd: The nucleophile was prepared in liquid ammonia
as before from Me3SnCl (0.5 mmol). Both substrates were
dissolved in 1 mL of anhydrous diethyl ether, and after 75 min
of irradiation the reaction was quenched with NH4NO3 in
Ack n ow led gm en t. This work was supported in part
by the Consejo de Investigaciones de la Provincia de
Co´rdoba (CONICOR), the Consejo Nacional de Investi-
gaciones Cient´ıficas y Te´cnicas (CONICET) (Argentina),
Antorchas Foundation and Stiftung Volkswagenwerk
(Germany).
J O960965U