Synthesis, reactions, and structural and NMR features of [2.2]metacyclophane monoenes and their tricarbonylchromium and cyclopentadienyliron(+) complexes

Article abstract of DOI:10.1021/jo990721k

8,16-Dimethyl-, 5,8,13,16-tetramethyl-, and 4,6,8,12,14,16- hexamethyl[2.2]metacyclophanene have been synthesized from the corresponding methyl-substituted 3-thia[3.2]metacyclophane precursors via a Wittig rearrangement-Hofmann elimination procedure. Simp

Full text of DOI:10.1021/jo990721k

7140
J . Org. Chem. 1999, 64, 7140-7152
Syn th esis, Rea ction s, a n d Str u ctu r a l a n d NMR F ea tu r es of
[2.2]Meta cyclop h a n e Mon oen es a n d Th eir Tr ica r bon ylch r om iu m
a n d Cyclop en ta d ien ylir on (+) Com p lexes
Reginald H. Mitchell* and Limin Zhang
Department of Chemistry, University of Victoria, PO Box 3065, Victoria,
British Columbia, Canada V8W 3V6
Received April 29, 1999
8,16-Dimethyl-, 5,8,13,16-tetramethyl-, and 4,6,8,12,14,16-hexamethyl[2.2]metacyclophanene have
been synthesized from the corresponding methyl-substituted 3-thia[3.2]metacyclophane precursors
via a Wittig rearrangement-Hofmann elimination procedure. Simple addition of bromine or similar
electrophiles to the bridge double bond of the cyclophane monoenes did not occur; rather, the methyl-
substituted dihydropyrenes were formed. However, mono- and bis-tricarbonylchromium and mono-
cyclopentadienyliron complexes were obtained using ligand-exchange reactions. Addition of bromine
to the cyclophane bridge double bond of the iron complex did occur, but unusually slowly.
Surprisingly, debromination rather than dehydrobromination occurred when the dibromo addition
product was treated with a variety of bases. Photoisomerization of the monoenes and nucleophilic
substitution of the metal complexes was also investigated. The geometries of the monoenes and
their complexes were compared to the cyclophanes and the cyclophanedienes and to the monothia-
and dithiacyclophanes, by comparison of X-ray and calculated structural data and NMR spectro-
scopic data. Introduction of double bonds into the cyclophane bridges causes the cyclophane step
to be less steep but increases distortion of the internal atoms out of the plane of the benzene rings.
Making the bridges nonidentical also causes a sideways twist of the step.
In tr od u ction
tion of the unsaturated bridge using the thiacyclophane
5 as intermediate, and so was able to obtain the tert-
butyl-substituted examples 6 (R ) Me, Et). However, the
In contrast to the case of [2.2]paracyclophane,1 the
saturated bridges of [2.2]metacyclophane cannot be
directly functionalized.2 This led Boekelheide3 to devise
a rather lengthy synthesis of the cyclophane monoenes
1 (R ) H, CH3), in which the unsaturated bridge was
formed first via a Wittig reaction. Closure of the second
(saturated) bridge occurred by means of a Wurtz cou-
pling, which proceeded rather poorly (10%) for R ) CH3,
making study of the chemistry of the cyclophane mo-
noenes difficult. Our synthesis4 of metacyclophanedienes,
2, using thiacyclophane intermediates gave much im-
proved access to both these compounds and cyclophanes
with unsaturated bridges in general.5 However, study of
the chemistry of [2.2]metacyclophane-1,9-dienes with
internal alkyl substituents, 2, was still not possible
because of the thermal valence isomerization that readily
occurred to the more stable and much more reactive
dihydropyrenes 3.6 Tashiro7 recognized that the thiacy-
clophane route might yield better quantities of the
metacyclophane monoenes by first forming the saturated
bridge to form a 1,2-diphenylethane, 4, and then forma-
* To whom correspondence should be addressed. Phone: 250-721-
7159. Fax: 250-721-7147. E-mail: regmitch@uvic.ca.
(1) Dewhirst, K. C.; Cram, D. J . J . Am. Chem. Soc. 1958, 80, 3115.
(2) Lindsay, W. S.; Stokes, P.; Humber, L. G.; Boekelheide, V. J .
Am. Chem. Soc. 1961, 83, 943.
(3) Blaschke, H.; Ramey, C. E.; Calder, I.; Boekelheide, V. J . Am.
Chem. Soc. 1970, 92, 3675.
(4) Mitchell, R. H.; Boekelheide, V. J . Am. Chem. Soc. 1974, 96,
1547.
(5) Mitchell, R. H. Heterocycles 1978, 11, 563.
tert-butyl substituents substantially change the chemis-
try of the [2.2]metacyclophanes, relative to those cyclo-
phanes that lack such groups,7,8 and thus, we thought
the chemistry of the monoenes without such groups
should also be explored, and as well we wanted to study
some of their metal complexes, to complement our studies
in cyclophanes with saturated bridges.9
(6) Phillips, J . B.; Molyneux, R. J .; Sturm, E.; Boekelheide, V. J .
Am. Chem. Soc. 1967, 89, 1704; Mitchell, R. H. In Advances in
Theoretically Interesting Molececules; Thummel, R. P., Ed.; J AI Press:
New York, 1989; Vol. 1, p 135. Mitchell, R. H.; Iyer, V. S.; Mahadevan,
R.; Venugopalan, S.; Zhou, P. J . Org. Chem. 1996, 61, 5116.
(7) Tashiro, M.; Yamato, T. J . Am. Chem. Soc. 1982, 104, 3701.
Tashiro, M.; Yamato, T.; Kobayashi, K. J . Org. Chem. 1984, 49, 4724.
(8) Tashiro, M.; Yamato, T. J . Org. Chem. 1981, 46, 1544. Tashiro,
M.; Mataka, S.; Takezaki, Y.; Takeshita, M.; Arimura, T.; Tsuge, A.;
Yamoto, T. J . Org. Chem. 1989, 54, 451. Yamato, T.; Ide, S.; Tokuhisa,
K.; Tashiro, M. J . Org. Chem. 1992, 57, 271.
(9) Mitchell, R. H.; Vinod, T. K.; Bushnell, G. W. J . Am. Chem. Soc.
1990, 112, 3487.
10.1021/jo990721k CCC: $18.00 © 1999 American Chemical Society
Published on Web 08/31/1999

Features of [2.2]Metacyclophane Monoenes
J . Org. Chem., Vol. 64, No. 19, 1999 7141
Syn th eses
Since we knew we would want to investigate metal
complexation of the cyclophane monoenes, we decided
that 7a -c would be appropriate targets, with different
numbers of methyl substituents and, hence, different
electron density in the benzene rings. To make use of the
thiacyclophane intermediate, 8a -c, we thus required the
three diarylethanes 9a -c. Unfortunately, each of these
yield via diazotization.12 The latter failed to yield the
alcohol 20 via the Grignard reagent and formaldehyde
and so had to be put through a sequence, 18 f 21, similar
to that used in Scheme 1, to give the bromide 21. To
required their own route. Only 9c was attainable fairly
directly: bromomethylation of mesitylene proceeded in
85% yield to form 10 when myristyltrimethylammonium
bromide was used as phase-transfer catalyst10 with
trioxane and concentrated aqueous HBr at 80-90 °C.
Coupling of 10 with Mg in refluxing THF was almost
quantitative to give 11, which on rebromomethylation as
above, using 2 equiv of trioxane, gave 85% of the desired
product 9c and none of the isomer with both bromom-
ethyl groups in the same benzene ring. All compounds
achieve benzylic coupling of 21 without aryl-coupling
interference, Mg in THF in the presence of FeCl3 was
used2 and gave the dimer 22A in 93% yield. Repeat of
this four- step sequence on 22A then yielded 9b in 71%
yield.
Th ia cyclop h a n es 8. Thiacyclophanes can be obtained
by coupling one molecule containing a bromide and one
a thiol4 or by coupling a dibromide and Na2S‚9H2O.13 To
obtain good yields in the latter coupling, the dibromide
was placed in one dropping funnel and the sulfide in a
second, and the drop rates from both funnels were kept
as equal as possible. Typically, yields were 20-50%.
Bodwell14 has introduced sodium sulfide adsorbed on to
alumina to improve yields in this reaction and has
obtained yields as high as 65% in intermolecular cycliza-
tions using this reagent. We took a slightly different
approach. The dibromides 9 are not very soluble in
benzene. We thus thought that by keeping most of the
dibromide present as a suspended solid only a small
amount would be dissolved to couple, and thus, we would
achieve a high dilution process without the need for large
volumes of solvent. After a few trials, a suitable solvent
system was found for each of the dibromides 9a -c, shown
in Table 1, which gave very respectable yields of the
thiacyclophanes 8.
were fully characterized, and details are in the Experi-
mental Section. To access 9a , commercial 2,6-dichloro-
toluene was converted to the bromide 12 in 79% yield in
two steps11 via the mono-Grignard reagent and formal-
dehyde, followed by HBr, and then 12 could be coupled
in 88% yield to the dimer 13 using Mg in THF at room
temperature, which avoided formation of the aryl Grig-
nard and its subsequent coupling to byproducts. Dichlo-
ride 13 was converted to 9a in 66% overall yield using
the longer sequence shown in Scheme 1, rather than the
shorter Grignard-formaldehyde-HBr sequence used
above for 12, which from 13 only gave 9a in 25-35%
overall yield.
In each case, only the anti isomers were obtained, in
which the internal methyl groups are shielded to δ 0.8
(from about 2.2) by the opposite benzene rings.
Th e Meta cyclop h a n e Mon oen es 7. Wittig rear-
rangement15 of the three thiacyclophanes 8 proceeded
smoothly with n-BuLi or LDA in THF at room temper-
To access 9b was more difficult: commercial 2,5-
dimethylaniline was brominated in 94% yield in aqueous
alcohol to give 17A, which was deaminated to 17B in 75%
Sch em e 1
(10) Mitchell, R. H.; Iyer, V. S. Synlett 1989, 55.
(11) Mitchell, R. H.; Lai, Y. H. J . Org. Chem. 1984, 49, 2534.
(12) From an adaption of a procedure of: Organic Syntheses;
Wiley: New York, 1943; Collect. Vol. II, p 592.
(13) Mitchell, R. H.; Boekelheide, V. Tetrahedron. Lett. 1970, 11,
1197. Boekelheide, V.; Hollins, R. A. J . Am. Chem. Soc. 1973, 95, 3201.
Mitchell, R. H.; Yan, J . S.; Dingle, T. W. J . Am. Chem. Soc. 1982, 104,
2551.
(14) Bodwell, G. J .; Houghton, T. J .; Koury, H. E.; Yarlagadda, B.
Synlett 1995, 751.

7142 J . Org. Chem., Vol. 64, No. 19, 1999
Mitchell and Zhang
Sch em e 2
F igu r e 1. Equatorial and axial isomers of cyclophanes 23
(only one R group shown for clarity).
Ta ble 1. Solven t System Used to Cou p le Dibr om id es 9
w ith Sod iu m Su lfid e
bromide product yield (%) solvent (water/ethanol/benzene)
9a
9b
9c
8a
8b
8c
66
83
72
23:77:0
16:70:14
12:63:25
ature, followed by MeI, and gave essentially quantitative
yields of the [2.2]cyclophanes 23 as a mixture of equato-
rial and axial isomers, Figure 1.
stable compound, and both the paracyclophanyne 2618
and the metaparacyclophanyne 2719 can be trapped with
furan.
In the case of 23a , the ratio of equatorial to axial
isomers obtained was 91:9. These could be separated by
chromatography and showed mp 95-96 °C and 156-157
°C, respectively. They are easily distinguished by 1H
NMR, since the internal methyl protons of a x-23a appear
as two well-separated singlets at δ 0.53 and 0.88, the
latter being deshielded by the adjacent -SMe group,
while in eq-23a they appear at δ 0.57 and 0.59. In this
case, the adjacent ring hydrogen, R in Figure 1, is
deshielded to δ 7.70, while in a x-23a it appears at δ 7.26.
For 23b, the ratio of isomers was also 91:9, but for 23c
the ratio was 72:28. Less equatorial isomer was formed
in this case, presumably because of the unfavorable steric
interaction between the eq-Me and the ring-Me (R ) Me
in Figure 1). We were not able to separate these two
isomers by chromatography; however, we did obtain a
pure sample of eq-23c, because it reacts slower than the
axial isomer in the next step (below). Methylation of
23a -c with dimethoxycarbonium fluoroborate16 gave the
sulfonium salts 24 in high yield, which with t-BuOK in
t-BuOH/DMF (1:1) at room temperature underwent Hof-
mann elimination to give the desired monoenes 7a -c in
92, 72, and 52% yields, respectively. In the last case, 26%
of eq-23c was returned. Evidently, the additional ring
Me groups in eq-23c (R ) Me in Figure 1) inhibit
approach of the base to remove the â-eq-H in the syn
transition state.
We thus first tried addition of bromine to the double
bonds of the monoenes 7. Even with a solution of bromine
in dichloromethane at -78 °C, the monoenes 7a -c
immediately gave dark green solutions of the dihydro-
pyrenes 28a -c in essentially quantitative yield. The
same result was obtained using the milder brominating
reagents pyridinium bromide perbromide20 and NBS21
also in dichloromethane at 0 °C. Excess bromination
reagents introduced further bromine substituents into
23
the dihydropyrene ring. Both PhSeCl22 and PhSeCl3
also gave the dihydropyrenes, rather than addition
product. Evidently, the monoenes 7 are extremely easily
oxidized to the dihydropyrenes 28. Because on addition
of bromine, the green dihydropyrene color forms im-
mediately, a reasonable mechanism would involve after
attack by the electrophile, a very rapid loss of H+ with
formation of the interannular bond, such that a final loss
of HBr gives the aromatic dihydropyrene. One possible
mechanism is shown in Scheme 2.
Tashiro7 reported the reaction of the monoene 6
(R)Me,Et) only with excess bromine in carbon tetrachlo-
ride, and found the tetrabrominated dihydropyrene,
4,5,9,10-tetrabromo-28 (R ) H; R′ ) t-Bu) to be the
product. With chlorine, a hexachloro derivative was
obtained, while iodine chloride gave a tetrachloropyrene
in which the internal alkyl groups were lost.7 Even when
nonoxidizing electrophiles, such as CF3COOH, were used,
By this route, 1-2 g of 7a ,b and 10 g of 7c could be
obtained in a single run compared to the 90 mg of 7a
prepared in ref 3, such that the chemistry of the mo-
noenes 7 can now be explored.
E lect r op h ilic Ad d it ion R ea ct ion s of t h e Mo-
n oen es 7. One of the potentially more important syn-
thetic utilities of the monoenes 7 would be to be precur-
sors to the as yet unknown [2.2]metacyclophanynes, e.g.,
24. The [2.2]orthocyclophanyne 25 is known17 and is a
(17) Meier, H.; Gugel, H. Synthesis 1976, 338.
(18) Psiorz, M.; Hopf, H. Angew. Chem., Int. Ed. Engl. 1982, 21,
623.
(19) Chan, C. W.; Wong, H. N. J . Am. Chem. Soc. 1988, 110, 462.
(20) Djerassi, C.; Scholtz, C. R. J . Am. Chem. Soc. 1948, 70, 417.
(21) Mitchell, R. H.; Lai, Y. H.; Williams, R. V. J . Org. Chem. 1979,
44, 4733. Mitchell, R. H.; Chen, Y.; Zhang, Org. Prep. Proc. Int. 1997,
29, 715.
(15) Mitchell, R. H.; Otsubo, T.; Boekelheide, V. Tetrahedron. Lett.
1975, 16, 219.
(16) Borch, R. F. J . Org. Chem. 1969, 34, 627.
(22) Sharpless, K. B.; Lauer, R. F. J . Org. Chem. 1974, 39, 429.
(23) Engman, L. J . Org. Chem. 1987, 52, 4086.

Features of [2.2]Metacyclophane Monoenes
J . Org. Chem., Vol. 64, No. 19, 1999 7143
interannular bond formation to form a dihydropyrene
was observed.24 To overcome this problem, we thought
constraining the π-electrons of the benzene ring as a
metal complex might be useful and would make inter-
annular bond formation less likely.
Met a l Com p lex F or m a t ion of t h e Mon oen es 7.
Two metal complexes were chosen for study: the tricar-
bonylchromium derivatives and the cyclopentadienyl iron
salts. The former was chosen both for comparison with
our previously studied thia-25 and saturated [2.2]cyclo-
phane9 examples, and as well as candidates for nucleo-
philic and electrophilic substitution; the latter was chosen
to be subjected to bromination studies, since the cationic
iron center is not removed by bromine as is a tricarbo-
nylchromium group.
dichloromethane in the dark was not instantaneous. In
the case of 32b, after 10 min, the expected dibromide 34b
was formed. This was easily verified by the disappear-
ance of the vinylic proton doublets (J ) 11.1 Hz) of 32b
at δ 7.08 and 6.49 and the appearance of two new sets of
doublets (J ) 5.0 Hz) at δ 5.57 and 5.23 for the -CHBr-
protons of 34b. A single trans isomer appeared to form.
Since the internal methyl groups of 34b were slightly
deshielded relative to those in 32b, while the aryl
hydrogens remained almost unchanged, the two bromines
probably occupy the pseudoaxial positions, adjacent to
the internal methyl groups as shown. The other hydro-
gens and carbons, though slightly shifted, all remained.
In the case of 32c however, not only did the vinylic proton
doublets disappear but also the uncomplexed aromatic
ring proton, and the product was the tribromide 35c. As
well, the 13C NMR signals for the vinylic carbons of 32b
at δ 138.2 and 131.8 were replaced by -CHBr- carbons
for 34b at δ 54.2 and 51.1. In the unsubstituted case,
32a , about equal amounts of the dibromide 34a and the
tribromide 35a were obtained. Clearly, addition of bro-
mine to the vinylic double bond is much slower for the
complexed compounds than for the uncomplexed, such
that ring bromination competes. We have previously
reported9 that the bis-Cr(CO)3-complexed derivative of
syn-2 (R ) H) could not be catalytically hydrogenated,
even though uncomplexed examples reduce rapidly.
Taken together, this suggests that metal complexation
of the metacyclophanenes substantially reduces the
available electron density in the bridge double bond, even
though the π-orbital overlap between the aromatic ring
and the bridge double bond must only be partial. Sur-
prisingly when the tribromide 35c was treated with base,
(NaOMe/MeOH or KOBu-t/HOBu-t/THF or DBU or
NaNH2/THF), debromination and decomplexation oc-
curred to give the monoene 36 and none of the vinyl
bromide and none of the dihydropyrene.
Reaction of the monoenes 7a ,c with a 3:1 molar excess
of tricarbonylchromiumnaphthalene26 in ether at 60 °C
in a sealed tube gave the mono- and bis-chromium
complexes 29a ,c and 30a ,c, respectively. The less sub-
stituted monoene 7a gave a 2:3 ratio of mono/bis com-
plexes 29a and 30a in 95% overall yield, while the more
substituted monoene gave a 1:4 ratio of 29c/30c in 82%
overall yield. Their structures were given by their clear
molecular ions in their mass spectra and by their proton
and carbon NMR spectra. In the bis-complexes the
aromatic ring protons were all shielded to δ 5.1-5.4,
characteristic9,25 of such protons, while in the mono
complexes, one set was shielded and one set remained
relatively normal at around δ 7; their spectra are
discussed in detail below.
Several CpFe+ complexes of [2.2]metacyclophanes were
successfully prepared by Swann and Boekelheide,27 using
the photochemical exchange procedure of Gill and Mann,28
in which η6-p-xylene-η5-cyclopentadienyliron hexafluoro-
phosphate, 31, is used as the CpFe+ source. Indeed,
irradiation of a 2:1 mixture of 31 and monoene 7c in dry
dichloromethane using a garden flood lamp gave the
monoiron complex 32c in 82% yield. However, using the
less electron-rich monoene 7a failed to exchange the
p-xylene ligand of 31. Thus, the p-chlorotoluene contain-
ing reagent 33 was used in place of 31, and with monoene
7a gave 78% of the iron complex 32a . With monoene 7c
it gave the same yield of 32c that 31 did. Interestingly,
the yield with monoene 7b was lower in both cases, 47%
with 33 and 43% with 31. There structures were proven
by their NMR spectra (discussed below) and elemental
analyses, and as well, an X-ray structure was obtained
for 32a .
Formylation of the complex 32a with TiCl4 and Cl2-
CHOMe proceeded poorly and gave only 2% of the
aldehyde 37 while returning 32% of unchanged starting
material.
Nu cleop h ilic Su bstitu tion of 29a . It is well-known
that reaction of n-BuLi with η6-tricarbonylchromiumben-
zene abstracts a ring proton to give a species that reacts
with electrophiles to yield a substituted benzene.29 Reac-
Br om in a tion of th e Ir on Com p lexes 32. Surpris-
ingly, reaction of the iron complexes 32 with bromine in
(24) Tashiro, M.; Kobayashi, K.; Yamato, T. Chem. Lett. 1985, 327.
(25) Mitchell, R. H.; Vinod, T. K.; Bodwell, G. J .; Bushnell, G. W. J .
Org. Chem. 1989, 54, 5871.
(26) Addition of 5% THF improved the reaction. Desobry, V.;
Kunding, E. P. Helv. Chim. Acta 1981, 64, 1288.
(27) Swann, R. T.; Boekelheide, V. J . Organomet. Chem. 1982, 231,
143.
(29) Semmelhack, M. F.; Hall, H. T.; Yoshifuji, M.; Clark, G. J . Am.
Chem. Soc. 1975, 97, 1247. Semmelhack, M. F.; Hall, H. T.; Farina,
R.; Yoshifuji, M.; Clark, G.; Bargar, T.; Hirotsu, K.; Clardy, J . J . Am.
Chem. Soc. 1979, 101, 3535.
(28) Gill, T. P.; Mann, K. R. Inorg. Chem. 1980, 19, 3007.

7144 J . Org. Chem., Vol. 64, No. 19, 1999
Mitchell and Zhang
tion of 29a with n-BuLi in THF at -78 °C and subse-
quent quenching with D2O gave an 87% yield of a mixture
of the ring monodeuterium-substituted complexes 38a ,b
together with a minor amount of the vinyl substituted
complex 38c. No benzylic substitution at the saturated
on the ring, ortho to the bridge, causes a pronounced
shielding of the bridge carbon of about 5 ppm per adjacent
methyl group,32 and indeed the bridge carbons, C-9,10,
of 7c are shielded by 6 ppm from those of 7a and 7b
(Table 2). The symmetry is maintained in the bis-
chromium complexes, 30a ,c, and again the axial hydro-
gens and bridge carbons of 30c are shielded from those
in 30a (Table 2). Interestingly, the chemical shifts of all
(etheno and ethano) bridge hydrogens (compare 30c and
7c, for example) are hardly changed on bis-complexation.
This suggests that there are several effects that cancel
each other out. In general, a bridge hydrogen of 30 will
be subject to a deshielding by the ring current of the
adjacent benzene ring, a shielding by the ring current of
the opposite benzene ring, and an anisotropic effect
caused by the chromium tricarbonyl group on the adja-
cent benzene ring (the opposite one is too far away).33
Since tricarbonylchromium complexation in effect re-
duces the ring current of the ring that is complexed,25,34
and both the adjacent and opposite rings are complexed,
then on bis-complexation, both the shielding and the
deshielding of the benzene rings change (from that in 7),
together with the new through space anisotropic effect
of the closer Cr(CO)3. Evidently in total, these just
balance. This is not true for the internal methyl groups,
Me-8 and Me-16, which are approximately 0.5 ppm (H)
and 7 ppm (C) less shielded in the complexes than the
parents. In general,35 the proton and carbon chemical
shifts of any substituent methyl groups of arenes are
hardly changed (<1 ppm) upon complexation. This is
indeed found to be so for the other (C-4,6) methyl groups
of 30c and so suggests that for the internal methyl groups
the dominant effect is the reduction in shielding by the
opposite benzene ring, with the other two effects canceling
each other. The monocomplexes 29a ,c can then be
understood. H-2 and both H-9 protons of 29a ,c are
assigned the shielded ones of each pair, by about 0.3-
1.0 ppm, again primarily because the opposite uncom-
plexed ring causes full shielding of these. Again, this is
especially evident for the internal methyl groups, where
C-8 is 0.6-0.8 ppm shielded from C-16, while the external
methyl groups, Me-6,12 are hardly affected. The three
iron complexes, 32a -c show very similar results (Table
2). For substituent methyl or methylene groups, com-
plexation shifts in such iron complexes35 are small and
very similar to those of the chromium complexes; how-
ever, the shielding effect on the aromatic ring carbons is
larger for iron complexes. The results for 32c and 29c
are consistent; for example, the chemical shifts of C-5
are δ 85.5 and 91.2, respectively.
ethano bridge occurred, unlike nonmetacyclophane cases.30
This was easily shown from the 2H NMR spectrum, where
the only peaks seen were at δ 5.43 and 5.24 for the new
deuterium atoms on the complexed ring and a very small
peak at δ 6.37 for the vinyl-deuterium. Mass spectroscopy
indicated only one deuterium was incorporated. No
selectivity between the ring hydrogens was found. Oxida-
tion of the intermediate anionic species with FeCl3/DMF
gave equal amounts of uncomplexed monoene 7a and
saturated cyclophane 39. When t-BuLi in THF at -78
°C was used, addition29 to the ring occurred, and then
with FeCl3/DMF, 62% of the tert-butyl substituted cyclo-
phanene 40 was obtained.
Geom etr y a n d NMR F ea tu r es of th e
Cyclop h a n en es a n d Th eir Meta l Com p lexes
The numbering system used is shown in Figure 2.
F igu r e 2. Numbering system used for the cyclophanenes and
their metal complexes.
The C2 rotation axis passing through the midpoints of
the C1-C2 and C9-C10 bonds for each of 7a -c makes
protons H-1 and H-2 magnetically equivalent, and they
thus appear as singlets at about δ 6.6 ((0.06 ppm), Table
2. The bridge protons however, are not equivalent and
show a typical31 AA′XX′ multiplet, which can be analyzed
with two chemical shifts and four coupling constants. The
values were refined by simulation, and the iterated
values are reported in Table 2.
The axial protons (δ 2.5) are more shielded than the
equatorial protons (δ 2.9), presumably by the ring current
of the opposite ring; however, in 7c, they are even more
so, δ 2.2, possibly because of additional twisting caused
by the C-6(Me)-H-9e steric interaction, not present in
7a ,b. This is supported also by a small (0.1 ppm) increase
in shielding of the internal methyl (Me-8,16) protons in
7c relative to those in 7a ,b. As well, there is a small
change in the coupling constants in 7c from those of 7a
and 7b, which are almost identical. In 13C NMR spectra
of [2.2]metacyclophanes, substitution of a methyl group
Ca lcu la ted a n d Exp er im en ta l Str u ctu r a l
F ea tu r es
The molecules shown in Figure 3 present an interest-
ing and relevant series to compare structural features.
The X-ray structure of the [2.2]metacyclophane 39 has
been reported by Hanson.36 We have obtained the X-ray
(32) Takemura, T.; Sato, T. Can. J . Chem. 1976, 54, 3412.
(33) Mitchell, R. H.; Chen, Y.; Khalifa, N.; Zhou, P. J . Am. Chem.
Soc. 1998, 120, 1785.
(34) Swann, R. T.; Boekelheide, V. J . Organomet. Chem. 1982, 231,
143. Langer, E.; Lehner, H. J . Organomet. Chem. 1979, 173, 47.
(35) Mori, N.; Takamori, M. J . Chem. Soc., Dalton Trans. 1985, 1661.
Segard, C.; Roques, B.; Pommier, C.; Guiochon, G. Anal. Chem. 1971,
43, 1146.
(30) Cara, R. J .; Trahanovsky, W. S. Tetrahedron. Lett. 1973, 14,
3823.
(31) Gunther, H. Angew. Chem., Int. Ed. Engl. 1972, 11, 861.
(36) Hanson, A. W. Acta Crystallogr. 1962, 15, 956.

Features of [2.2]Metacyclophane Monoenes
J . Org. Chem., Vol. 64, No. 19, 1999 7145
13C NMR (δ)
Ta ble 2. Selected NMR Da ta for th e Mon oen es a n d Th eir Com p lexes
1H NMR
J H-H (Hz)
J 9e-9a J 9e-10a
J 9a-10a
δ(H-1)
δ(H-2)
δ(H-9eq
)
δ(H-9ax
δ(H-10ax
)
)
C-9
C-8(Me)
compd
δ(H-10eq
)
J 9e-10e
δ(Int-CH3)
C-10
C-16(Me)
Ext-CH3
7a
6.59
6.54
6.64
2.89
2.50
-12.2
2.8
3.9
12.1
3.9
12.2
3.0
0.76
39.2
39.0
33.0
17.3
17.0
17.8
7b
7c
2.81
3.06
2.48
2.17
-12.1
0.73
0.63
20.6
2.8
-13.1
4.0
19.16
19.13
12.0
30a
30c
29a
6.65
6.69
6.68
6.35
6.95
6.36
7.13
6.53
7.08
6.49
7.24
6.58
2.97
3.12
2.77
3.07
2.94
3.20
3.00
3.11
2.95
3.05
3.17
3.25
2.52
2.17
2.19
2.77
1.92
2.43
2.42
2.58
2.42
2.54
2.18
2.18
1.20
1.11
0.51
1.31
0.49
1.11
0.67
0.92
0.62
0.90
0.49
0.86
37.7
32.5
39.4
24.2
24.4
20.5
19.0
-13.4
3.9
29c
32a
32b
32c
-13.0
3.1
3.1
12.7
3.6
12.3
3.4
33.2
32.1
38.4
37.4
38.2
37.0
32.6
32.4
21.1
21.0
20.9
22.3
20.7
22.1
21.1
22.5
19.1
18.9
-13.1
4.1
-12.9
3.9
20.4
20.2
19.2
19.1
12.0
F igu r e 4. Structural features referred to in Table 3. θ ) angle
that the internal benzene carbon (+) is bent out of the plane
of the benzene carbons (*). d1 ) distance (Å) the internal
methyl carbon projects over the opposite benzene ring; 0.0
would be over the opposite internal benzene carbon, 1.4 would
be over the center of the ring. d2 ) distance (Å) from the
internal methyl carbon to the plane (* carbons) of the opposite
benzene ring. d3 ) distance (Å) between internal methyl
carbons. d4 ) distance (Å) between internal benzene carbons.
F igu r e 3. Molecules selected for comparison.
structures of 8a , 41, and 32a , and they are deposited as
Supporting Information. Unfortunately, as yet, we have
not successfully determined the structures of 7a , 40, or
29a . Nevertheless, we have carried out PCMODEL37 (an
MM2 + π) and AM138 calculations on these molecules,
so that comparisons can be made between each other and
with the X-ray data. We have given both PCMODEL and
AM1 data because the latter do not contain the param-
eters for metals, and a comparison between the two is
thus useful to give confidence in the PCMODEL values
for metal containing systems.33 Selected data (see Figure
4 for an explanation) are compared in Table 3.
In our experience, calculations such as PCMODEL and
AM1 on the cyclophanes give geometries good enough for
discussion, though not necessarily agreeing with all the
X-ray-derived data. For example, in the [2.2]metacyclo-
phane 39 the bond lengths of bonds in or attached to the
benzene rings all agree very well (<0.01 Å), whereas the
bond length between the two bridge carbons is calculated
at 1.549 Å but found to be 1.573 Å. Since the π-π
compression between the rings is hard to model, one
might expect the biggest differences between calculation
and experiment to be in the distances and angles between
opposite rings, and in the distortion of the benzene rings
out of the plane. Some of the possible values are pre-
sented in Table 3 for discussion. For 39, it can be seen
that the experimental values for d4, d3, and d2 lie between
Ta ble 3. Com p a r ison of Ca lcu la ted a n d Exp er im en ta l
Va lu es for Selected Str u ctu r a l F ea tu r es of th e Molecu les
in F igu r e 3
compd
39
7a
40
8a
41
29a
32a
δ(int-C1H3) 0.56
0.76
1.52
0.85
1.30
0.51/
1.31
20.5/
20.9
0.7
0.67/
0.92
20.9/
22.3
δ(int-13CH3) 14.8
17.3
20.0
14.3
14.7
d1 (PCM)
d1 (AM1)
d1 (exp)
d2 (PCM)
d2 (AM1)
d2 (exp)
1.0
1.0
1.0
3.4
2.9
3.1
0.4
0.7
0.0
0.4
0.8
1.0
1.1
3.5
3.2
3.2
0.75
1.3
1.0
3.5
3.5
3.1
0.6/0.8
3.0
2.9
2.8
2.7
3.0
3.1
3.0
d3 (PCM)
d3 (AM1)
d3 (exp)
4.792 4.336 4.158 4.709 4.709 4.353 4.345
4.243 4.175 4.063 4.373 4.718
4.394
4.325 4.429
4.226
d4 (PCM)
d4 (AM1)
d4 (exp)
3.058 2.851 2.753 3.217 3.446 2.867 2.849
2.681 2.703 2.660 3.049 3.371
2.819
14.8
3.014 3.243
2.687
17.1/
13.4
θ (PCM)
16.3
11.0
18.9
12.0
10.9
7.3
16.5/
14.9
θ (AM1)
θ (exp)
9.4
8.7
7.1
3.6
4.6
3.4
10.3/10.4
the PCMODEL and AM1 values. PCMODEL appears to
overestimate the repulsion between the rings, while AM1
underestimates it. To decide whether the calculated data
can be used for cases in which X-ray data is lacking,
independent data are required. The NMR data in Table
3 provide some: From calculations of d1, it is predicted
in all cases that the internal methyl carbon (e.g., on C-16)
lies within the perimeter of the carbons of the opposite
(37) PCMODEL v. 5.13 for WINDOWS was used from Serena
Software, Box 3076, Bloomington, IN, 47402-3076.
(38) Hyperchem v 4.5; Hypercube, Inc., 419 Phillip St., Waterloo,
Ontario, Canada, N2L 3X2.

7146 J . Org. Chem., Vol. 64, No. 19, 1999
Mitchell and Zhang
F igu r e 5. Sketch of the relative geometries of 39, 7a , and
40.
F igu r e 6. Sketch of the axial alignments A-C.
Ta ble 4. Th er m od yn a m ic P a r a m eter s for th e Th er m a l
Retu r n of 42 to 7 in Aceton itr ile
benzene ring, being closest to the center of the opposite
ring for 39 (d1 ) 1) and closest to the edge (C-8) for 40
(d1 ∼ 0). We would expect the chemical shift of the
internal carbon (and its protons) to decrease (i.e., become
more shielded) as d1 goes from 0 to 1.4 and the carbon
approaches the center of the ring current of the opposite
∆Hf(42-7)
a
rate at 25 °C
Eact
∆HqThinSpacea ∆SqThinSpaceb
[AM1]
compd (103k, min-1) (kcal/mol) (kcal/mol)
(cal/K)
(kcal/mol)
42a 52.5 ((0.8)
21.3
22.2
24.6
20.7
21.6
24.0
-14.3
-12.8
-18.7
7.33
7.00
7.63
42b
42c
6.42 ((0.1)
2.04 ((0.1)
1
ring. The observed 13C and H shifts for 40, 7a , and 39
clearly are in the correct order to that expected for the
calculated values of d1. These chemical shift shieldings
also depend on d2, the distance from the internal carbon
to the plane of the opposite benzene ring. However, in
these three compounds this is reasonably constant at 3
Å. Using the J ohnson-Bovey tables,39 the predicted
shieldings caused by the opposite benzene’s ring current
at this value for d2 are from about 1.9 ppm over the center
of the ring to 0.7 ppm over its edge, similar to the range
found for the protons in Table 3. Note: an aryl methyl is
normally at δ 2.3. From Table 3, θ, the bending of the
internal benzene carbon out of the plane of the rest of
the benzene carbons is calculated most poorly. Clearly
PCMODEL allows the benzene to bend out of plane too
much, as does AM1, though to a lesser extent. This leads
to larger distances for d3 and d4 in PCMODEL than in
AM1.
a
b
Estimated error is (0.5 kcal/mol. Estimated error is (2 cal/
K.
displaced; see C in Figure 6. The sideways displacement
follows the order 41 > 8a > 7a ≈ 32a ≈ 29a. The twist
follows the order 29a ≈ 7a > 32a > 8a .
P h otova len ce Isom er iza tion of th e Mon oen es
7a -c
In the same way that 2 and 3 are valence isomers that
are photochemically reversibly interconvertible, and as
well thermally from 2 to 3,6,40 the monoenes 7a -c can
be converted by irradiation with light of wavelength 254
nm to their photoisomers 42a -c, which thermally revert
Taking all of the data together, it suggests that on
progression from the cyclophane 39 to the monoene 7a
to the diene 40, the geometry of the molecule changes
such that the two benzene rings both slide away from
each other along the long axis of the molecule (the C-5-
C-13 distance goes from 6.488 to 6.828 to 7.063 Å, and
d1 decreases) but move so that the distance between the
planes of the rings decreases from 2.4 to 2.2 to 1.8 Å,
such that as a result d3 and d4 decrease and θ increases.
We have attempted to sketch this in Figure 5.
For the metal complexes 29a and 32a , there appears
to be very little geometry change on complexation from
that of the monoene 7a . In fact, PCMODEL predicts the
geometry of 32a quite well compared to its X-ray struc-
ture, with the exception of the overestimate of the
bending of the benzene ring out of plane. From calcula-
tion and X-ray data, the sulfides 8a and 41 appear to
have geometry closest to 39 of the three shown in Figure
5. This is consistent with the chemical shift data for the
internal methyl carbon atoms.
Finally, the relative orientation of the rings to the C5-
C13 axis is of interest: for 39 both X-ray and calculations
have the two benzene rings parallel, i.e., projected on to
a plane, carbons 13-16-8-5 form a straight line,
likewise for 14-15-3-4 and 12-11-7-6, see A in
Figure 6. The diene 40 is calculated to be the same. For
the disulfide 41, both X-ray and calculations have the
rings parallel as for 39, but shifted sideways off the axis;
see B in Figure 6. All of the remaining molecules have
the rings twisted off the C5-C13 axis as well as sideways
to the monoenes on standing. The thermal return reac-
tion has been studied for 1 (where R ) H, H; H, Me; Me,
Me),41 where it was found that substitution of the internal
hydrogens by methyl groups speeded the thermal return
reaction. Study of 7a -c would show the effect of external
methyl groups. The thermal return of 42 to 7 in aceto-
nitrile was thus followed by UV spectroscopy at six
temperatures between 15 and 40 °C to obtain the
thermodynamic parameters of Table 4.
As can be seen, the rates of thermal return to the
cyclophane monoene follow the order 42a > 42b > 42c,
and thus, external methyl groups slow the rate of thermal
return. Blattmann and Schmidt40 found that for the
thermal return of 2 f 3(R ) Me), four similarly substi-
tuted external methyl groups, raised ∆Hq and Eact by
about 1.5 kcal/mol and changed ∆Sq by 5 cal/K, consistent
with our results, but overall did not change the rates
much. From the calculated ∆Hf values, as expected, the
cyclophane monoenes are the more stable, but adding
external methyl groups does not favor thermally one or
other isomer much at all, and thus, perhaps the main
factor reducing the rate with increasing substitution is
in the organization involved to get to the transition
state.42
Con clu sion s
This study of three cyclophane monoenes and two of
their metal complexes has made clear their structural
(40) Blattmann, H. R.; Schmidt, W. Tetrahedron 1970, 26, 5885.
(41) Ramey, C. E.; Boekelheide, V. J . Am. Chem. Soc. 1970, 92, 3681.
(42) Muszkat, K. A.; Fischer, E. J . Chem. Soc. B 1967, 662.
(39) Emsley, J . W.; Feeney, J .; Sutcliffe, L. H. High-resolution NMR
spectroscopy; Pergamon: New York, 1965; Vol. VI, Appendix B.

Features of [2.2]Metacyclophane Monoenes
J . Org. Chem., Vol. 64, No. 19, 1999 7147
under reflux for 16 h. The organic layer was separated,
washed, and dried by azeotropic distillation, which was
continued to remove most of the benzene. The resulting
bromide 12 was diluted with dry THF (150 mL), and a portion
(5 mL) was added to Mg (2.43 g, 0.1 mol) in dry THF (200
mL) under N2. The reaction was initiated by warming, but then
the rest of the bromide was added at such a rate to maintain
the reaction mixture close to room temperature (not reflux).
The mixture was then stirred for a further 6 h, and then water
and dichloromethane were added. The organic layer was
washed, dried, and evaporated to yield 26.1 g (95%) of the
dichloride 13 as a white powder, suitable for use in the next
step. A portion was recrystallized from hexane as white
crystals: mp 108-109 °C; 1H NMR (250 MHz) δ 7.24 (dd, J )
7.5, 1.7 Hz, 2), 7.05 (t, J ) 7.5 Hz, 2), 6.99 (dd, J ) 7.5, 1.7
Hz, 2), 2.87 (s, 4), 2.36 (s, 6); 13C NMR (62.9 MHz) δ 141.6,
135.1, 133.9, 127.6, 127.3, 126.6, 35.1, 15.8; EI MS m/z 278
(M+). Anal. Calcd for C16H16Cl2: C, 68.83; H, 5.78. Found: C,
68.71; H, 5.61.
1,2-Bis(3′-cya n o-2′-m eth ylp h en yl)eth a n e (14). Copper-
(I) cyanide (20.0 g, 0.26 mol) was added to a hot (∼100 °C)
solution of the dichloride 13 (27.9 g, 0.1 mol) in N-methyl-2-
pyrrolidinone (70 mL), and the mixture was heated to reflux
with mechanical stirring for 10 h. It was then cooled to ∼100
°C, a further portion of CuCN (13 g, 0.17 mol) was added, and
the mixture was then stirred and refluxed for a further 14 h.
This was then cooled to ∼100 °C, and a solution of ethylene-
diamine (50 mL) in water (100 mL) was added, with vigorous
stirring, which was continued for 15 min. The precipitated
solid was collected by filtration, was washed thoroughly with
ethylenediamine solution and then water, and then was
extracted with dichloromethane (800 mL). The extract was
washed, dried, and evaporated to give crude dinitrile. This was
preabsorbed on to silica gel and chromatographed on SiGel
using dichloromethane as eluant to yield 21.9 g (84%) of the
dinitrile 14 as a white solid. A portion was recrystallized from
hexane as white crystals: mp 204-205 °C; 1H NMR (360 MHz)
δ 7.48 (dd, J ) 7.4, 1.8 Hz, 2), 7.22 (dd, J ) 7.4, 1.8 Hz, 2),
7.18 (t, J ) 7.4 Hz, 2), 2.89 (s, 4) 2.48 (s, 6); 13C NMR (90.6
MHz) δ 140.4, 139.6, 133.4, 131.0, 126.5, 118.4, 113.7, 33.7,
17.3; IR (KBr) 2224 cm-1; CI MS m/z 261 (MH+). Anal. Calcd
for C18H16N2: C, 83.05; H, 6.19; N, 10.76. Found: C, 82.98; H,
6.20; N, 10.69.
1,2-Bis(3′-for m yl-2′-m eth ylp h en yl)eth a n e (15). Diisobu-
tylaluminum hydride (0.25 mol in hexane 200 mL) was added
dropwise under N2 with good stirring to a solution of the
dinitrile 14 (26.0 g, 0.1 mol) in dry benzene (350 mL) at 20
°C. The stirring was continued overnight, and then the mixture
was cooled in an ice bath and decomposed by slow addition of
methanol (frothing) (100 mL) and then aqueous HCl (400 mL,
1:1). The mixture was then extracted with dichloromethane
(∼ 1 L), and the organic layer was washed, dried, and
evaporated. The residue was preabsorbed and filtered through
a column (6 × 15 cm) of SiGel using 1:1 PE/dichloromethane
as eluant to give 22.3 g (84%) of the dialdehyde 15 as a pale
yellow powder. A small sample was recrystallized from acetone-
hexane: mp 116-117 °C; 1H NMR (360 MHz) δ 10.29(s, 2),
7.67 (dd, J ) 6.5, 2.7 Hz, 2), 7.30-7.26 (m, 4), 2.94 (s, 4), 2.61
(s, 6); 13C NMR (90.6 MHz) δ 192.9, 141.0, 138.2, 134.6, 134.7,
130.5, 125.9, 33.8, 14.0; IR (KBr) 2768, 1688 cm-1; CI MS m/z
267 (MH+). Anal. Calcd for C18H18O2: C, 81.17; H, 6.81.
Found: C, 81.32; H, 6.89.
1,2-Bis(3′-h yd r oxym eth yl-2′-m eth ylp h en yl)eth a n e (16).
A solution of the dialdehyde 15 (43.6 g, 164 mmol) in THF
(450 mL) was added dropwise to a stirred suspension of NaBH4
(4 g, 100 mmol) in THF (150 mL) and water (10 mL) at 20 °C.
After the suspension was stirred for 20 h, aqueousaqueous HCl
(60 mL of concentrated acid + 120 mL water) was added slowly
to decompose the reaction mixture. Filtration then gave the
first batch of product. The aqueous layer was extracted with
dichloromethane until it became clear, and then the combined
organic extracts were dried and evaporated to give a second
batch of product. The two batches were combined and dried
under vacuum to yield 43.5 g (98%) of the diol 16, suitable for
use in the next step, as a white powder: mp 156-158 °C (lit.2
relationship to the saturated cyclophanes and the cyclo-
phanedienes, and the mono- and dithiacyclophanes, by
detailed examination of NMR and X-ray data and calcu-
lation results. As well, we have shown that some explora-
tion of their chemical reactions is possible, unlike the
cyclophanedienes, which too easily thermally convert to
the dihydropyrenes, or the saturated cyclophanes, which
have essentially inert bridges.
Exp er im en ta l Section
Melting points were determined on a Reichert 7905 melting
point apparatus integrated to a chrome-alumel thermocouple.
UV-vis spectra were recorded in acetonitrile. Proton and
carbon NMR spectra were recorded in CDCl3 as solvent. Mass
spectra were recorded either using methane gas for chemical
ionization (CI) or using electron impact (EI) at 70 eV. Exact
mass measurements used perfluorokerosene as calibrant.
Elemental analyses were carried out by Canadian Microana-
lytical Services Ltd, Vancouver, BC. All evaporations were
carried out under reduced pressure on a rotary evaporator,
and all organic extracts were washed with water and dried
over anhydrous MgSO4. SiGel refers to Merck silica gel, 70-
230 mesh. PE refers to distilled petroleum ether, bp 30-60
°C.
2-Br om om eth yl-1,3,5-tr im eth ylben zen e, (10). A mixture
of mesitylene (120 g, 1 mol), trioxane (30 g, 0.33 mol), aqueous
HBr (48%, 500 mL), and myrystyltrimethylammonium bro-
mide (5 g, 15 mmol) was heated at 80-90 °C with vigorous
mechanical stirring for 8 h. PE (50 mL) was then carefully
added, and the mixture was left open to cool overnight without
further stirring. The solid that collected on top was collected
by filtration, washed with water, and air dried. Recrystalli-
zation from PE gave bromide 10 as white crystals: mp 49-50
1
°C (lit.43 mp 49-50 °C); H NMR (250 MHz) δ 6.84 (br s, 2),
4.55 (s, 2), 2.36 (s, 6), 2.25 (s, 3); 13C NMR (62.9 MHz) δ 138.4,
137.4, 131.0, 129.2, 29.6, 21.0, 19.1.
1,2-Bis(2′,4′,6′-tr im eth ylp h en yl)eth a n e (11). A portion
(about 5 mL) of a solution of bromide 10 (106 g, 0.5 mol) in
dry THF (400 mL) was added to Mg (6.03 g, 0.25 mol) in dry
THF (500 mL) under N2 with gentle warming to initiate the
reaction, and then the remainder of the bromide solution was
added at such a rate to keep the mixture at reflux. After the
addition, the reaction mixture was heated at reflux for a
further 2 h, and then the solvent was evaporated, the residual
solid was extracted with ether (800 mL) and water, and the
ether layer was washed well with water, dried, and evaporated
to give 66 g (99%) of dimer 11 as a fluffy white solid, suitable
for use in the next step. A portion was recrystallized from
hexane: mp 113-114 °C; 1H NMR (250 MHz) δ 6.86 (s, 4),
2.77 (s, 4), 2.36 (s, 12), 2.26 (s, 6); 13C NMR (62.9 MHz) δ 136.3,
135.7, 135.2, 129.1, 29.0, 20.8, 20.1; EI MS m/z 266 (M+). Anal.
Calcd for C20H26: C, 90.16; H, 9.84. Found: C, 90.26; H, 9.67.
1,2-Bis(3′-b r om om et h yl-2′,4′,6′-t r im et h ylp h en yl)et h -
a n e (9c). A mixture of dimer 11 (133 g, 0.5 mol), trioxane (33
g, 0.37 mol), aqueous HBr (48%, 300 mL), myrystyltrimethy-
lammonium bromide (4 g), and heptane (250 mL) were heated
at reflux with vigorous mechanical stirring for 20 h. After
cooling, the solid was collected by filtration, washed well with
water, and then dried to give 193 g (85%) of the bromide 9c,
suitable for use in the next step. A portion was recrystallized
from benzene as white crystals: mp 244-246 °C; 1H NMR (360
MHz) δ 6.85 (s, 2), 4.57 (s, 4), 2.80 (s, 4), 2.35 (s, 6), 2.33 (s, 6),
2.26 (s, 6); 13C NMR (90.6 MHz) δ 137.3, 136.9, 136.1, 135.1,
132.1, 130.6, 30.9, 29.4, 20.5, 19.2, 15.1; CI MS m/z 451 (MH+).
Anal. Calcd for C22H28Br2: C, 58.43; H, 6.24. Found: C, 58.50;
H, 6.03.
1,2-Bis(3′-ch lor o-2′-m eth ylp h en yl)eth a n e (13). A mix-
ture of 2-chloro-6-hydroxymethyltoluene11 (30.2 g, 0.2 mol) in
benzene (150 mL) and aqueous HBr (48%, 60 mL) was stirred
(43) Hauser, C. R.; Van Eenam, D. N. J . Am. Chem. Soc. 1957, 79,
5512.

7148 J . Org. Chem., Vol. 64, No. 19, 1999
Mitchell and Zhang
mp 158-160 °C); 1H NMR (360 MHz) δ 7.26-7.15 (m, 6), 4.72
(d, J ) 4 Hz, 4), 2.89 (s, 4), 2.30 (s, 6), 1.58 (bs, 2); 13C NMR
(90.6 MHz) δ 140.7, 139.0, 134.4, 129.0, 126.2, 125.9, 64.3, 34.8,
g, 50 mmol) in dry benzene (60 mL) at 20 °C. After the mixture
was stirred for 14 h, methanol (15 mL) was slowly added
followed by aqueous methanol (1:1, 40 mL)and aqueous 10%
HCl (120 mL). The mixture was filtered, the organic layer was
separated, and the aqueous layer was further extracted with
ether (2 × 100 mL). The combined organic layers were washed,
dried, and evaporated to give 10.33 g (97%) of the aldehyde
19 as a yellowish liquid suitable for use in the next step.
Vacuum distillation gave colorless liquid: bp 116-117°/2.5
mmHg;
14.1; IR (KBr) 3320, 3240 cm-1
.
1,2-Bis(3′-br om om eth yl-2′-m eth ylp h en yl)eth a n e (9a ).
A suspension of the diol 16 (8.78 g, 32.5 mmol) in benzene
(100 mL) and concentrated aqueous HBr (48%, 60 mL) and
concentrated H2SO4 (0.5 mL) was refluxed for 24 h. The
organic layer was separated, and the aqueous layer was
extracted with dichloromethane. The combined organic layers
were washed well with water, aqueous NaHCO3, and water
and then were dried and evaporated. The residue (CAU-
TION: lacrimator) was filtered through a column of SiGel (5
× 15 cm) using PE/dichloromethane (5:1) as eluant to yield
11.95 g (93%) of dibromide 9a as a white solid: mp 140-142
°C (lit.2 mp 142-143 °C); 1H NMR (250 MHz) δ 7.24-7.09 (m,
6), 4.54 (s, 4), 2.88 (s, 4), 2.30(s, 6); 13C NMR (90.6 MHz) δ
140.9, 136.1, 135.3, 130.1, 128.3, 126.1, 34.7, 33.4, 14.3.
2,6-Dibr om o-p-xylen e (17B). This procedure is adapted
from a similar one in Organic Syntheses.12 A rapid stream of
air saturated with bromine vapor (500 g, 3.12 mol) was pulled
by suction through a well stirred solution of 2,5-dimethyla-
niline (189 g, 1.56 mol) in 95% aqueous ethanol (1.7 L) and
water (1.4 L) at 20 °C in the dark (∼6 h). The solid 2,4-dibromo-
3,5-dimethylaniline44 (17A) was collected by filtration, washed
well with water, and used directly in the next step. A small
quantity was dried, indicated a yield of 94%, and gave 1H NMR
(90 MHz) δ 7.22 (s, 1), 3.37 (bs, 2), 2.51 (s, 3), 2.15 (s, 3). The
damp crude dibromide was dissolved in benzene (400 mL) and
95% aqueous ethanol (1.7 L), and then concentrated H2SO4
(200 mL) was added with stirring. NaNO2 (200 g) was then
added to this hot mixture in small portions as quickly as the
violence of the reaction would permit (∼1 h). After the addition,
the mixture was refluxed for 3 h and then was kept at ∼50 °C
for 14 h. The mixture was then cooled, and concentrated H2-
SO4 (200 mL) that had been diluted with water (1.4 L) was
added followed by PE (1.5 L). The organic layer was washed
with water, aqueous NaHCO3, and water and then was dried
and evaporated. The residue was distilled under vacuum (133-
136 °C/13 Torr) to give 289 g (75%) of the dibromide 17B: mp
1H NMR (250 MHz) δ 10.19 (s, 1), 7.59 (s, 1), 7.54 (s,
1), 2.66 (s, 3), 2.34 (s, 3); IR (film) 2730, 1690 cm-1; CI MS m/z
215, 213 (MH+, Br1).
1-Br om o-2,5-d im eth yl-3-h yd r oxym eth ylben zen e (20).
A solution of the aldehyde 19 (30.0 g, 140 mmol) in THF (not
dried, 350 mL) was added dropwise to a stirred suspension of
NaBH4 (2.65 g, 70 mmol) in THF (300 mL) at 20 °C. After the
mixture was stirred for 20 h, aqueous 10% HCl (180 mL) was
added, followed by ether (250 mL). The organic layer was
separated, and the aqueous layer was re-extracted with ether
(2 × 200 mL). The combined organic layers were washed, dried,
and evaporated to give 30.3 g (100%) of the alcohol 20 as a
1
white solid: mp 89-90 °C; H NMR (360 MHz) δ 7.33 (s, 1),
7.11 (s, 1), 4.67 (d, J ) 5.7 Hz, 2), 2.36 (s, 3), 2.28 (s, 3), 1.54
(t, J ) 5.7 Hz, 1); 13C NMR (90.6 MHz) δ 140.1, 137.0, 132.5,
132.4, 127.7, 125.9, 64.1, 20.5, 17.9; IR (KBr) 3240 (w) cm-1
;
EI MS m/z 216, 214 (M+, small, Br1). Anal. Calcd for C9H11
-
BrO: C, 50.26; H, 5.15. Found: C, 50.20; H, 5.15.
1-Br om o-3-br om om eth yl-2,5-d im eth ylben zen e (21). A
mixture of the alcohol 20 (28.8 g, 134 mmol), benzene (120
mL), and concentrated aqueous 48% HBr (100 mL) was
refluxed with stirring for 40 h. The organic layer was sepa-
rated, washed with water, aqueous NaHCO3, and water, and
then dried and evaporated. The residue was preabsorbed and
filtered through a short column of SiGel using PE as the eluant
and gave 36.3 g (97%) of the bromide 21 as a white solid. A
portion was recrystallized from pentane as colorless crystals:
mp 51-52 °C; 1H NMR (360 MHz) δ 7.34 (s, 1), 7.05 (s, 1),
4.47 (s, 2), 2.41 (s, 3), 2.26 (s, 3); 13C NMR (90.6 MHz) δ 137.2,
137.1, 133.8, 133.7, 130.0, 126.2, 32.8, 20.4, 18.3; CI MS m/z
281, 279, 277 (MH+, Br2). Anal. Calcd for C9H10Br2: C, 38.89;
H, 3.63. Found: C, 39.02; H, 3.62.
1,2-Bis-(3′-br om o-2′,5′-d im eth ylp h en yl)eth a n e (22A). A
solution of dibromide 21 (21.5 g, 7.7 mmol) in dry ether (50
mL) was warmed (to reflux if necessary) with Mg (0.95 g, 3.9
mmol) under N2. Once the reaction initiated, anhyd FeCl3 (50
mg, 0.3 mmol) was added, which caused vigorous reaction that
subsided after about 20 min. The reaction was then refluxed
for 1 h, and then cold water was added, followed by aqueous
HCl and dichloromethane. The organic layer was washed,
dried, and evaporated to give 14.2 g (93%) of the diphenyle-
thane 22A as a white powder. A portion was recrystallized
from ethyl acetate as white crystals: mp 169-170 °C; 1H NMR
(360 MHz) δ 7.26 (br s, 2), 6.85 (br s, 2), 2.79 (s, 4), 2.34 (s, 6),
2.25 (s, 6); 13C NMR (90.6 MHz) δ 141.4, 136.9, 132.3, 131.1,
129.1, 125.9, 35.7, 20.5, 18.6; CI MS m/z 399, 397, 395 (MH+,
Br2). Anal. Calcd for C18H20Br2: C, 54.57; H, 5.09. Found: C,
54.83; H, 5.06.
1,2-Bis(3′-cya n o-2′,5′-d im eth ylp h en yl)eth a n e (22B). Cu-
(I)CN (26.9 g, 0.3 mol) was added to a hot solution of the
dibromide 22A (39.6 g, 0.1 mol) in N-methyl-2-pyrrolidinone
(150 mL), and the mixture was mechanically stirred under
reflux for 20 h. The mixture was then cooled to 80 °C, aqueous
25% ethylenediamine (280 mL) was added, and the mixture
was vigorously stirred for 30 min. The solid was collected by
filtration, and the filtrate was extracted with dichloromethane
(2 × 100 mL). The solid was dissolved in dichloromethane
(∼800 mL) and was filtered. The organic layers were combined,
washed, dried, and evaporated. The residue was preabsorbed
and filtered through a column of SiGel using PE/dichlo-
romethane (1:1) as eluant to give 27.6 g, (96%) of the dinitrile
22B as a yellowish solid. A portion was recrystallized from
acetone as white crystals: mp 206-207 °C; 1H NMR (360 MHz)
δ 7.28 (s, 2), 7.05 (s, 2), 2.80 (s, 4), 2.43 (s, 6), 2.28 (s, 6); 13C
NMR (90.6 MHz) δ 140.4, 136.5, 136.3, 134.4, 131.1, 118.0,
113.5, 33.9, 20.6, 16.9; IR (KBr) 2223 cm-1; CI MS m/z 289
1
32-34 °C (lit.42a,b mp 32,36 °C); H NMR (90 MHz) δ 7.32 (s,
2), 2.47 (s, 3), 2.22 (s, 3).
3-Br om o-2,5-d im eth ylben zon itr ile (18). Cu(I)CN (29.5
g, 0.33 mol) was added to a mechanically stirred, hot (∼80 °C)
solution of 2,6-dibromo-p-xylene (17B) (174 g, 0.66 mol) in
DMF (100 mL). The mixture was then stirred at reflux for 4
h, and while still hot a solution of ethylenediamine (80 mL)
and water (240 mL) was added and, after some cooling,
dichloromethane (400 mL). After the mixture was stirred for
30 min, insoluble material was filtered, and the organic layer
was separated, washed, dried, and evaporated. The residue
was preabsorbed and chromatographed on SiGel using first
PE as eluant to elute 92 g (53%) of unreacted 17B. Elution
with PE/dichloromethane (3:1) then gave 61 g (44%) of the
mononitrile 18 as a white solid: mp 74-75 °C; 1H NMR (360
MHz) δ 7.57 (s, 1), 7.35 (s, 1), 2.56 (s, 3), 2.31 (s, 3); 13C NMR
(90.6 MHz) δ 138.4, 137.9, 137.6, 132.1, 125.5, 117.6, 114.0,
21.1, 20.3; IR (KBr) 2230 cm-1; CI MS m/z 212,210 (1:1, MH+).
Anal. Calcd for C9H8BrN: C, 51.46; H, 3.84; N, 6.67. Found:
C, 51.10; H, 3.81; N, 6.51.
Eluted last was 2.6 g (3%) of 2,5-dimethyl-1,3-phthalodini-
1
trile: mp 171-172 °C; H NMR (360 MHz) δ 7.61 (s, 2), 2.69
13
(s, 3), 2.38 (s, 3); C NMR (90.6 MHz) δ 142.4, 137.6, 136.9,
116.4, 114.5, 20.3, 18.9; IR (KBr) 2220 cm-1; CI MS m/z 157
(MH+).
Use of a full equivalent of CuCN gave 16-17% unreacted
dibromide, 50% of nitrile 18, and 20% of dinitrile, but the
reaction mixture was more difficult to separate.
3-Br om o-2,5-d im eth ylben za ld eh yd e (19). Diisobutyla-
luminum hydride (75 mmol in hexane (60 mL)) was added
dropwise over 1.5 h under N2 to a solution of nitrile 18 (10.5
(44) (a) Blanksma, J . J . Chem. Weekblad, 10, 136-141; Chem. Abstr.
1913, 7, 1493. (b) Bures, E.; Meskan, F. Casopis Ceskoslov. Lekarnicta
1937, 17, 149; Chem. Abstr. 1937, 31, 7857.

Features of [2.2]Metacyclophane Monoenes
J . Org. Chem., Vol. 64, No. 19, 1999 7149
(MH+). Anal. Calcd for C20H20N2: C, 83.30; H, 6.99; N, 9.71.
Found: C, 83.23; H, 7.16; N, 9.56.
details of the X-ray determination can be found in the
Supporting Information.
1,2-Bis(3′-for m yl-2′,5′-dim eth ylph en yl)eth an e (22C). Di-
isobutylaluminum hydride (0.25 mmol in hexane (200 mL))
was added dropwise under N2 to a solution of the dinitrile 22B
(28.84 g, 0.1 mol) in dry benzene (500 mL) at 20 °C. The
mixture was then stirred for 14 h and then decomposed
cautiously with methanol (100 mL) and aqueous HCl (1:1, 400
mL). The mixture was extracted with dichloromethane (800
mL), washing the solids well with dichloromethane, and then
the organic layers were combined, washed, dried, and evapo-
rated. The residue was preabsorbed and chromatographed on
SiGel using PE/CH2Cl2/ethyl acetate (5.5:4:0.5) as eluant to
give 24.2 g (82%) of the dialdehyde 22C as a yellowish solid.
A portion was recrystallized from acetone as white crystals:
mp 142-144 °C; 1H NMR (360 MHz) δ 10.28 (s, 2), 7.49 (s, 2),
7.13 (s, 2), 2.86 (s, 4), 2.57 (s, 6), 2.33 (s, 6); 13C NMR (90.6
MHz) δ 193.3, 141.2, 135.7, 135.6, 135.3, 134.6, 130.9, 34.1,
20.7, 13.7; IR (KBr) 2766, 1670 cm-1; CI MS m/z 295 (MH+).
Anal. Calcd for C20H22O2: C, 81.60; H, 7.53. Found: C, 81.44;
H, 7.68.
a n t i -6,9,14,17-T e t r a m e t h y l-2-t h ia [3.2]m e t a c y c lo -
p h a n e (8b). A solution of Na2S‚9H2O (2.64 g, 11 mmol) in N2-
purged water (60 mL), aqueous 95% ethanol (250 mL), and
benzene (50 mL) was added dropwise over 20 h to a well-
stirred suspension of the dibromide 9b (4.24 g, 10 mmol) and
Cs2CO3 (0.65 g) in N2-purged benzene (150 mL), aqueous 95%
ethanol (800 mL), and water (140 mL) under N2. Stirring was
continued for a further 2 h, and then the solvents were
evaporated. Dichloromethane and water were added to the
resultant solids, and the organic layer was separated, washed,
dried, and evaporated. The residue was preabsorbed on and
chromatographed over SiGel (4 × 40 cm) using PE/CH2Cl2 (4:
1) as eluant to yield 2.46 g (83%) of the thiacyclophane 8b. A
portion was recrystallized from hexane as white crystals: mp
1
175-176 °C; H NMR (360 MHz) δ 6.915 (s, 2), 6.905 (s, 2),
3.80 (AB, J ) 13.8 Hz, 2, Heq), 3.71 (AB, J ) 13.8 Hz, 2, Hax),
2.85 (AA′XX′, J AA′ ) 2.7 Hz, J AX ) -12.5 Hz, J AX′ ) 4.4 Hz, 2,
Heq), 2.60 (AA′XX′, J XX′ ) 12.5 Hz, 2, Hax), 2.24 (s, 6), 0.84 (s,
6); 13C NMR (90.6 MHz) δ 137.6, 136.7, 134.9, 134.5, 129.8,
129.3, 35.5, 34.2, 20.6, 13.9; CI MS m/z 297 (MH+). Anal. Calcd
for C20H24S: C, 81.03; H, 8.16. Found: C, 81.13; H, 8.02.
a n t i -5,7,9,13,15,17-H e x a m e t h y l-2-t h i a [3.2]m e t a -
cyclop h a n e (8c). A solution of Na2S‚9H2O (26.4 g, 0.11 mol)
in N2-purged water (440 mL) to which aqueous 95% ethanol
(760 mL) was added was then added dropwise over about 48
h to a well stirred suspension of the dibromide 9c (45.2 g, 0.10
mol) in N2-purged benzene (560 mL), aqueous 95% ethanol
(1425 mL), and water (200 mL) under N2. Stirring was
continued for a further 4 h, and then the solvents were
evaporated. Dichloromethane (1200 mL) and water (600 mL)
were added to the resultant solids, and the organic layer was
separated, washed, dried, and evaporated. The residue was
preabsorbed on and filtered through SiGel (12 × 30 cm) using
PE/CH2Cl2 (3:1) as eluant to yield 23.8 g (72%) of the
thiacyclophane 8c. A portion was recrystallized from hexane
1,2-Bis(3′-h yd r oxym e t h yl-2′,5′d im e t h ylp h e n yl)e t h -
a n e (22D). A solution of the dialdehyde 22C (21.0 g, 71.3
mmol) in THF (not dried, 500 mL) was added slowly to a
stirred suspension of NaBH4 (1.90 g, 50 mmol) in THF (75 mL)
and water (5 mL) at 20 °C. After 24 h, the mixture was
decomposed by addition of aqueous HCl (10%, 90 mL), and
the first batch of product was collected by filtration. A further
batch was obtained by extracting the aqueous layer with ether
(3 × 200 mL), which was then washed, dried, and evaporated.
The combined batches of diol 22D were dried and gave 20.2 g
(95%) of white solid. A portion was recrystallized from acetone
1
as colorless crystals: mp 174-175 °C; H NMR (360 MHz) δ
7.04 (s, 2), 6.98 (s, 2), 4.68 (d, J ) 5.7 Hz, 4), 2.81 (s, 4), 2.30
(s, 6), 2.27 (s, 6), 1.48 (t, J ) 5.7 Hz, 2); IR (KBr) 3280 (br)
cm-1. Anal. Calcd for C20H26O2: C, 80.50; H, 8.78. Found: C,
79.86; H, 8.24.
1
1,2-Bis(3′-b r om om et h yl-2′,5′-d im et h ylp h en yl)et h a n e
(9b). A mixture of the diol 22D (19.0 g, 64 mmol), benzene
(200 mL), and concentrated aqueous HBr (48%, 120 mL) was
refluxed for 36 h. On cooling, some product crystallized, which
was collected. The organic layer was washed, dried, and
evaporated, and the combined solids were preabsorbed and
filtered through a short column of SiGel using PE/CH2Cl2 (2:
1) as eluant to give 25.7 g (95%) of dibromide 9b as a white
solid. A portion was recrystallized twice from acetone as white
as colorless crystals: mp 208-210 °C; H NMR (360 MHz) δ
6.73 (s, 2), 3.82 (AX, J ) 13.5 Hz, 2, Heq), 3.70 (AX, J ) 13.5
Hz, 2, Hax), 3.02 (AA′XX′, J AX ) -13.3 Hz, J AX′ ) 4.2 Hz, J AA′
) 3.0 Hz, 2, Heq), 2.51 (AA′XX′, J XX′ ) 12.8 Hz, 2, Hax), 2.39 (s,
6), 2.34 (s, 6), 0.81 (s, 6, int-Me); 13C NMR (90.6 MHz) δ 140.1,
135.0, 134.8, 133.1, 130.9, 129.2, 29.6, 27.9, 19.9, 19.7, 15.0;
CI MS m/z 325 (MH+). Anal. Calcd for C22H28S: C, 81.42; H,
8.70. Found: C, 81.09; H, 8.63.
a n t i -8,16-D i m e t h y l-1-m e t h y lt h i o [2.2]m e t a c y c lo -
p h a n e (23a ). n-Butyllithium (5 mmol in hexane (2 mL)) was
added via syringe to a stirred solution of thiacyclophane 8a
(1.07 g, 4.0 mmol) in dry THF (120 mL) under N2 at 20 °C.
After 5 min, methyl iodide (0.34 mL, 5.5 mmol) was added,
and then after a further 2 min, aqueous HCl and PE were
added. The organic layer was separated, washed, dried, and
evaporated to give a mixture of the two isomers of 23a as 1.14
g (100%) as a yellowish solid. For synthetic purposes, this
crude mixture could be used directly in the next step. To
separate the isomers, a small portion (130 mg) was chromato-
graphed over SiGel (2 × 80 cm) using PE as eluant. Eluted
first was 113 mg (91%) of eq-23a as a white solid: mp 95-96
°C; 1H NMR (COSY) (360 MHz) δ 7.70 (br d, J ) 7.3 Hz, H-14),
7.17 (d, J ) 7.4 Hz, H-12), 7.13 (2d, H-4,6), 6.97 (t, H-13), 6.84
(t, H-5), 4.02 (dd, J ) 11.3, 4.2 Hz, H-1ax), 3.18 (dd, J ) 12.5,
4.2 Hz, H-2eq), 2.98-2.94 (m, H-9eq,10eq), 2.76-2.70 (m, H-9ax,-
10ax), 2.64 (t, J ) ∼11.9 Hz, H-2ax), 2.12 (s, 3), 0.59 (s, 3), 0.57
(s, 3); 13C NMR (90.6 MHz) δ 143.0, 142.5, 137.2, 136.9, 135.6,
135.1, 128.1, 127.6, 127.3, 124.7, 124.4, 124.2, 43.2, 36.3, 36.1,
15.5, 14.9, 14.8; CI MS m/z 283 (MH+). Anal. Calcd for
1
crystals: mp 178-179 °C; H NMR (360 MHz) δ 7.03 (s, 2),
6.96 (s, 2), 4.51 (s, 4), 2.81 (s, 4), 2.28 (s, 12); 13C NMR (90.6
MHz) δ 141.0, 135.9, 135.5, 132.1, 130.9, 128.9, 34.9, 33.6, 20.8,
13.9. Anal. Calcd for C20H24Br2: C, 56.63; H, 5.70. Found: C,
56.64; H, 5.57.
a n ti-9,17-Dim eth yl-2-th ia [3.2]m eta cyclop h a n e (8a ). A
solution of Na2S‚9H2O (3.7 g, 15 mmol) in N2-purged water
(120 mL) and aqueous 95% ethanol (380 mL) was added
dropwise over 48 h to a well-stirred suspension of the dibro-
mide 9a (5.55 g, 14 mmol) in N2-purged aqueous 95% ethanol
(618 mL) and water (182 mL) under N2. Stirring was continued
for a further 2 h, and then the solvents were evaporated.
Dichloromethane and water were added to the resultant solids,
and the organic layer was separated, washed, dried, and
evaporated. The residue was preabsorbed on and chromato-
graphed over SiGel using PE/CH2Cl2 (3:1) as eluant to yield
2.32 g (66%) of the thiacyclophane 8a . A portion was recrystal-
lized from hexane as white crystals: mp 171-173 °C; 1H NMR
(360 MHz) δ 7.13-7.10 (m, 4), 6.98 (t, J ) 7.5 Hz, 2), 3.85
(AB, J ) 13.8 Hz, 2, Heq), 3.73 (AB, J ) 13.8 Hz, 2, Hax), 2.94
(AA′XX′, J AA′ ) 2.6 Hz, J AX ) -12.6 Hz, J AX′ ) 4.5 Hz, 2, Heq),
2.64 (AA′XX′, J XX′ ) 12.4 Hz, Hax), 0.85 (s, 6); 13C NMR (90.6
MHz) δ 139.7, 137.6, 135.1, 129.1, 128.7, 125.6, 35.6, 34.1, 14.3;
EI MS m/z 268 (M+). Anal. Calcd for C18H20S: C, 80.55; H,
7.51. Found: C, 79.73; 7.24.
C
19H22S: C, 80.80; H, 7.85. Found: C, 81.10; H, 7.76.
Eluted next was 11.5 mg (9%) of a x-23a as a white solid:
mp 156-157 °C; 1H NMR (360 MHz) δ 7.26 (d, J ) 7.2 Hz, 1),
7.16 (d, J ) 7.2 Hz, 1), 7.15 (d, J ) 7.2 Hz, 1), 7.07 (d, J ) 7.3
Hz, 1), 6.88 (t, J ) 7.3 Hz, 1), 6.83 (t, J ) 7.3 Hz, 1), 4.41 (br
d, J ) 5.8 Hz, H-1eq), 3.28 (dd, J ) 14.0, 6.9 Hz, H-2eq), 3.04
(br d, J ) 14.0 Hz, H-2ax), 2.96-2.81 (m, H-9eq,10eq), 2.78-
2.70 (m, H-9ax,10ax), 2.18 (s, 3), 0.88 (s, C-16(Me)), 0.53 (s,
C-8(Me)); 13C NMR (90.6 MHz) δ 144.3, 142.3, 137.1, 136.9,
An X-ray structure was obtained on a crystal 0.45 × 0.77 ×
0.68 mm in size and was found to be orthorhombic, space group
Pccn, with a ) 28.790 Å, b ) 12.922 Å, c ) 7.575 Å, and R )
â ) γ ) 90°. The structure was refined to R ) 0.082. Full

7150 J . Org. Chem., Vol. 64, No. 19, 1999
Mitchell and Zhang
134.7, 133.9, 128.8, 128.7, 128.5, 128.1, 124.0, 123.9, 57.6, 41.1,
36.2, 36.0, 17.2, 14.7, 14.4; CI MS m/z 283 (MH+).
h until clean powdery solid was obtained (this step can be
repeated if necessary), which was collected, 1.24 g (90%) of
sulfonium salt 24a , used directly in the elimination step below.
a n t i -1 -M e t h y l t h i o -5 ,8 ,1 3 ,1 6 -t e t r a m e t h y l [ 2 .2 ] -
m eta cyclop h a n e (23b). A solution of LDA (freshly prepared
from n-butyllithium (6 mmol in hexane (4 mL)) and i-Pr2NH
(0.85 mL, 6 mmol)) in dry THF (10 mL) was added via syringe
to a stirred solution of thiacyclophane 8b (1.19 g, 4 mmol) in
dry THF (120 mL) under N2 at 20 °C. After 30 min, methyl
iodide (0.40 mL, 6.5 mmol) was added, followed after 2 min
by aqueous HCl and diethyl ether. The organic layer was
separated, washed, dried, and evaporated. The residue was
preabsorbed and filtered through a column of SiGel using PE/
dichloromethane (4:1) as eluant and gave 1.24 g (100%) of the
two isomers (91:9) of cyclophane 23b as a semisolid. For
synthetic purposes, this was used in the next step. A portion
(100 mg) was chromatographed through a column (2 × 65 cm)
of SiGel, using PE/dichloromethane (10:1) as eluant. Eluted
first was 90 mg (91%) of eq-23b as a white crystals from
ether: mp 114-115 °C; 1H NMR (360 MHz) δ 7.47 (s, 1, H-14),
6.96, 6.93, 6.90 (s, 1 each), 3.99 (dd, J ) 11.3, 4.3 Hz, 1, H-1ax),
3.11 (dd, J ) 12.4, 4.2 Hz, 1, H-2eq), 2.90-2.85 (m, 2, H-9eq,-
10eq), 2.73-2.67 (m, 2, H-9ax,10ax), 2.61 (t J ) ∼11.9 Hz, 1,
H-2ax), 2.26 (s, 3), 2.20 (s, 3), 2.13 (s, 3), 0.60 (s, 3), 0.58 (s, 3);
13C NMR (90.6 MHz) δ 140.4, 139.9, 137.3, 137.0, 135.6, 135.1,
133.6, 133.2, 128.9, 128.3, 128.0, 124.8, 52.3, 43.2, 36.2, 36.0,
20.9, 20.6, 15.5, 14.7, 14.5; CI MS m/z 311 (MH+). Anal. Calcd
for C21H26S: C, 81.23; H, 8.44. Found: C, 81.32; H, 8.38.
Eluted next was 9 mg (9%) of a x-23b as a white solid: mp
176-177 °C; 1H NMR (360 MHz) δ 7.05 (s, 1, H-14), 6.94 (s, 2,
H-4,6), 6.86 (s, 1, H-12), 4.34 (dd, J ) 7.0, 0.8 Hz, 1, H-1eq),
3.26 (dd, J ) 14.0, 7.1 Hz, 1, H-2eq), 2.97 (dd, J ) 13.9, 0.8 Hz,
1, H-2ax), 2.88-2.69 (m, 4, H-9,10), 2.23 (s, 3), 2.20 (s, 3), 2.16
(s, 3), 0.89 (s, 3), 0.54(s, 3); 13C NMR (90.6 MHz) δ 141.7, 139.7,
137.0, 136.9, 134.6, 133.8, 132.79, 132.77, 129.5, 129.3, 129.2,
128.7, 57.5, 41.0, 36.2, 35.9, 20.7, 20.4, 17.1, 14.4, 14.1.
a n t i -4,6,8,12,14,16-H e x a m e t h y l-1-m e t h y lt h io [2.2]-
m eta cyclop h a n e (23c). A solution of LDA (freshly prepared
from n-butyllithium (48 mmol in hexane (32 mL)) and i-Pr2-
NH (7.0 mL, 50 mmol)) in dry THF (100 mL) was added over
30 min to a stirred solution of thiacyclophane 8c (13.0 g, 40
mmol) in dry THF (600 mL) under N2 at 20 °C. After 30 min,
methyl iodide (3.2 mL, 52 mmol) was added, followed after 2
min by aqueous HCl and diethyl ether (500 mL). The organic
layer was separated, washed, dried, and evaporated. The
residue was filtered through a short column of SiGel using
PE/dichloromethane (3:1) as eluant and gave 13.4 g (99%) of
the two isomers (72-eq/28-ax) of cyclophane 23c as a white
solid, CI MS m/z 339 (MH+). The ratio of the isomers was
obtained from the internal methyl proton integrations, a x-23c
at δ 0.79 and 0.41; eq-23c at δ 0.57 and 0.56. We were not
able to separate these isomers by chromatography. For syn-
thetic purposes, the mixed isomers were used directly in the
next step.
Likewise, from 23b (2.48 g, 8 mmol), Borch reagent (2.0 g,
80% oil, 10 mmol), dichloromethane (25 mL), and ethyl acetate
(40 mL) was obtained 24b, 3.28 g, (99%) and from 24c (13.5
g, 40 mmol), Borch reagent (10 g, 80% oil, 50 mmol), dichlo-
romethane (120 mL), and ethyl acetate (60 mL) was obtained
24c, 16.5 g, (94%).
a n ti-8,16-Dim eth yl[2.2]m eta cyclop h a n -1-en e (7a ). Po-
tassium tert-butoxide (0.62 g, 5.5 mmol) was added to a stirred
solution of the sulfonium salts 24a (1.10 g, 2.87 mmol) in DMF/
tert-butyl alcohol (1:1, 26 mL) at 20 °C under N2. After 30 min,
PE (300 mL) and water were added. The organic layer was
separated, washed, dried, and evaporated. The residue was
preabsorbed and filtered through a short column of SiGel using
PE as eluant and gave 0.62 g (92%) of monoene 7a as white
crystals from methanol-hexane: mp 148-150 °C (lit.3 mp
1
151-152 °C); H NMR (360 MHz) δ 7.00-6.95 (m, 4), 6.88-
6.84 (m, 2), 6.59 (s, 2), 2.89 (AA′XX, J AA′ ) 2.8, J AX ) -12.2,
J AX′ ) 3.9 Hz, 2, H-9eq,10eq), 2.50 (AA′XX′, J XX′ ) 12.1 Hz, 2,
H-9ax,10ax), 0.76 (s, 6); 13C NMR (90.6 MHz) δ 140.4, 138.5,
135.9, 132.1, 128.1, 125.8, 123.7, 39.2, 17.3; UV (CH3CN) λmax
(ꢀmax) nm 212 (36 300), 255 (24 300), 307 (1980); CI MS m/z
235 (MH+).
a n ti-5,8,13,16-Tet r a m et h yl[2.2]m et a cyclop h a n -1-en e
(7b). From potassium tert-butoxide (1.6 g, 14.3 mmol), sulfo-
nium salts 24b (2.35 g, 5.7 mmol), and DMF/t-BuOH (1:1, 50
mL), exactly as described for 7a above, there was obtained 1.08
g (72%) of monoene 7b as white crystals from methanol: mp
151-152 °C; 1H NMR (360 MHz) δ 6.77 (s, 2), 6.66 (s, 2), 6.54
(s, 2), 2.81 (AA′XX, J AA′ ) 2.8, J AX ) -12.1, J AX′ ) 3.9 Hz, 2,
H-9eq,10eq), 2.48 (AA′XX′, J XX′ ) 12.2 Hz, 2, H-9ax,10ax), 2.22
(s, 6), 0.73 (s, 6); 13C NMR (90.6 MHz) δ 138.4, 137.9, 135.9,
134.8, 132.1, 128.8, 124.2, 39.0, 20.6, 17.0; UV (CH3CN) λmax
(ꢀmax) nm 214 (37,000), 260 (28,400), 301 (1,930); CI MS m/z
263 (MH+). Anal. Calcd for C20H22: C, 91.55; H, 8.45. Found:
C, 91.42; H, 8.38.
a n ti-4,6,8,12,14,16-Hexa m et h yl[2.2]m et a cyclop h a n -1-
en e (7c). From potassium tert-butoxide (13.8 g, 124 mmol),
sulfonium salts 24c (27.2 g, 62 mmol), and DMF/t-BuOH (1:1,
400 mL), exactly as described for 7a above, there was obtained
9.36 g (52%) of monoene 7c, as white crystals from methanol:
1
mp 220 °C dec; H NMR (360 MHz) δ 6.68 (s, 2), 6.64 (s, 2),
3.06 (AA′XX′, J AA′ ) 4.0, J ax ) -13.1, J AX′ ) 3.0 Hz, 2, H-9eq,-
10eq), 2.28 (s, 6), 2.22 (s, 6) 2.17 (AA′XX′, J XX′ ) 12.0 Hz, 2,
H-9ax,10ax), 0.63 (s, 6); 13C NMR (90.6 MHz) δ 139.5, 135.3,
133.9, 132.6, 131.5, 129.8, 129.2, 33.0, 19.16, 19.13, 17.8; UV
(CH3CN) λmax (ꢀmax) nm 217 (44,600), 260 (29,100), ∼315 (2,-
400); CI MS m/z 291 (MH+). Anal. Calcd for C22H26: C, 90.98;
H, 9.02. Found: C, 90.91; H, 8.76.
Eluted after 7c from the column was the equatorial isomer
eq-24c, 5.3 g (26%), characterized above.
Some pure eq-23c was returned in the elimination step
below (to give monoene 7c) and thus could be characterized
Rea ction of th e Mon oen e 7c w ith Br om in e To Give
Dih yd r op yr en e 28c. A solution of Br2 (0.1 M in dichlo-
romethane, 10 mL) was added at -78 °C to a solution of
monoene 7c (290 mg, 1 mmol) in dichloromethane (10 mL).
The solution turned dark green immediately. Water was
added, and the organic layer was washed, dried and evapo-
rated and filtered through a short column of SiGel using PE
as eluant to give 280 mg (97%) of dihydropyrene 28c, identical
(TLC, NMR) to an authentic sample.45
1
as white plates from hexane/methanol: mp 142-143 °C; H
NMR (360 MHz) δ 6.58 (s, 1), 6.57 (s, 1), 3.68 (dd, J ) 11.4,
2.7 Hz, 1, H-1ax), 3.25 (dd, J ) 12.9, 2.7 Hz, 1, H-2eq), 3.10-
3.03 (m, 2, H-9eq,10eq), 2.85 (dd, J ) 12.7, 11.6 Hz, 1, H-2ax),
2.57 (s, 3), 2.46-2.36 (m, 2, H-9ax,10ax), 2.34 (s, 3), 2.33 (s, 3),
2.31 (s, 3), 2.25 (s, 3), 0.57 (s, 3), 0.56 (s, 3); 13C NMR (90.6
MHz) δ 142.4, 141.9, 134.6, 134.5, 133.6, 133.5, 132.7, 131.4,
130.5, 127.9, 52.0, 37.4, 29.9, 29.6, 21.1, 19.2, 17.2, 15.7.
By subtraction of these peaks from those of the mixed
isomers, the 1H NMR spectrum of a x-23c was indicated: δ
6.65 (s, 1), 6.64 (s, 1) 4.88 (dd, J ) 6.5, 1.0 Hz, H-1eq), 3.40
(dd, J ) 14.5, 1.4 Hz, H-2eq), 3.12-3.04 (m, H-9eq,10eq), 2.93
(dd, J ) 14.7, 6.9 Hz, H-2ax), 2.52 (s, 3), 2.44-2.37 (m, H-9ax,-
10ax), 2.36 (s, 3), 2.21 (s, 3), 0.79 (s, 3), 0.41 (s, 3).
1H NMR of 28c (360 MHz) δ 8.55 (s, 4), 7.79 (s, 2), 3.15 (s,
12), -4.02 (s, 6); 13C NMR (90.6 MHz) δ 134.0, 128.7, 118.7,
31.7, 19.8, 14.6.
Similar results were obtained from 7a to give 28a 4 and from
7b to give 28b.46
When pyridinium bromide perbromide (320 mg, 1 mmol)
instead of bromine was used, essentially the same amount of
product 28c was obtained.
Su lfon iu m Sa lts 24a -c. Dimethoxycarbonium fluorobo-
rate (Borch reagent)16 (0.88 g, 80% oil, 5.4 mmol) was added
to a stirred solution of mixed isomers of 23a (1.02 g, 3.6 mmol)
in dichloromethane (20 mL) at -30 °C and was stirred for 4 h
with warming to 20 °C. The solvent was then decanted, and
ethyl acetate (10 mL) was added and stirring continued for 4
(45) Renfroe, H. B., Gurney, J . A.; Hall, L. A. R. J . Org. Chem. 1972,
20, 3045.

Features of [2.2]Metacyclophane Monoenes
J . Org. Chem., Vol. 64, No. 19, 1999 7151
Tr ica r bon ylch r om iu m (0) Com p lexes 29c a n d 30c of
th e Mon oen e 7c. A solution of monoene 7c (87 mg, 0.3 mmol)
and tricarbonylchromiumnaphthalene26 (164 mg, 1 mmol) in
N2-purged ether (1 mL) and THF (0.2 mL) was stirred in a
small heavy-walled screw cap vial in an oil bath at 60 °C for
20 h. The insoluble bis-complex 30c was collected by filtration
and recrystallized from hexane/dichloromethane as an orange-
red solid, 111 mg (66%): mp 267-268 °C dec (sealed tube);
1H NMR (250 MHz) δ 6.69 (s, 2, H-1,2), 5.06 (s, 2, H-5,13),
∼3.18-3.06 (m, 2, H-9eq,10eq), 2.23 (s, 6), ∼2.22-2.11 (m, 2,
H-9ax,10ax), 2.13 (s, 6), 1.11 (s, 6); 13C NMR (62.9 MHz) δ 233.8,
130.4, 115.7, 113.2, 108.4, 107.2, 98.0, 90.7, 32.5, 24.4, 19.01,
19.00; IR (KBr) 1840-90 (br), 1930-60 (br) cm-1; CI MS m/z
563 (MH+). Anal. Calcd for C28H26Cr2O6: C, 59.79; H, 4.66.
Found: C, 59.57; H, 4.71.
The mother liquor was evaporated and the residue was
chromatographed on SiGel under N2 using PE/dichloromethane
as eluant to give 20 mg (16%) of the monocomplex 29c as a
red solid: mp 196-198 °C; 1H NMR (250 MHz) δ 6.95 (d, J )
11.3 Hz, 1), 6.72 (s, 1), 6.36 (d, J ) 11.3 Hz, 1), 5.10 (s, 1),
3.20, 2.94, 2.43, 1.92 (see Table 2, H-9,10), 2.26, 2.24, 2.22,
2.10 (s, 3 each), 1.11 (s, 3), 0.49 (s, 3); 13C NMR (62.9 MHz) δ
235.0, 138.9, 135.3, 134.3, 134.2, 133.2, 131.5, 130.4, 125.2,
116.8, 114.6, 107.3, 105.9, 100.5, 91.2, 33.2, 32.1, 21.1, 21.0,
19.11, 19.09, 18.90, 18.89; IR (KBr) 1860, 1870, 1945, 1955
cm-1; CI MS m/z 427 (MH+).
Tr ica r bon ylch r om iu m (0) Com p lexes 29a a n d 30a of
th e Mon oen e 7a . From monoene 7a (70 mg, 0.3 mmol) and
tricarbonylchromiumnaphthalene (130 mg, 0.8 mmol) exactly
as described above for 7c was obtained 87 mg (57%) of the
orange-red bis-complex 30a , mp >245 °C dec, and 42 mg (38%)
of the red mono-complex 29a , mp 178-179 °C. Bis-complex,
30a : 1H NMR (250 MHz) δ 6.65 (s, 2), 5.45-5.28 (m, 6), 3.03-
2.92 (m, 2, H-9eq,10eq), 2.57-2.47 (m, 2, H-9ax,10ax), 1.20 (s, 6);
13C NMR (90.6 MHz) δ 233.2, 131.8, 118.3, 97.2, 95.9, 93.0,
85.5, 37.7, 24.2; IR (KBr) 1855 (br), 1942, 1955 cm-1; CI MS
m/z 507 (MH+). Anal. Calcd for C24H18Cr2O6: C, 56.92; H, 3.58.
Found: C, 57.08; H, 3.53. Mono-complex, 29a : 1H NMR (360
MHz) δ 7.05-6.93 (m, 3), 6.88 (d, J ) 11.1 Hz, 1), 6.35 (d, J )
11.1 Hz, 1), 5.44-5.20 (m, 3), 3.07, 2.77, 2.77, 2.19 (see Table
2, H-9,10), 1.31 (s, 3), 0.51 (s, 3); 13C NMR (62.9 MHz) δ 234.4,
139.7, 137.3, 136.5, 134.6, 129.2, 126.6(×2), 125.0, 119.6, 118.0,
99.6, 94.9, 91.8, 86.5, 39.4, 37.1, 20.9, 20.5; IR (KBr) 1855 (br),
1936, 1953 cm-1; CI MS m/z 371 (MH+).
Monoene 7b (118 mg) yielded the brown complex 32b: 112
mg (47%); mp >220 °C dec; 1H NMR (360 MHz, CD2Cl2) δ 7.08
(d, J ) 11.1 Hz, 1), 6.95 (s, 1), 6.84 (s, 1), 6.49 (d, J ) 11.1 Hz,
1), 5.89 (s, 1), 5.61 (s, 1), 4.66 (s, 5), 3.05, 2.95, 2.54, 2.42 (see
Table 2, H-9,10), 2.64 (s, 3), 2.27 (s, 3), 0.90 (s, 3), 0.62 (s, 3);
13C NMR (90.6 MHz, CD2Cl2) δ 139.8, 137.9, 137.4, 137.0,
134.5, 131.0, 125.8, 124.5, 116.5, 114.1, 96.8, 87.5, 86.5, 82.2,
77.0, 38.2, 37.0, 22.1, 20.7, 20.4, 20.2. Anal. Calcd for C25H27F6-
FeP: C, 56.84; H, 5.15. Found: C, 56.52; H, 5.36.
Monoene 7c (130 mg) yielded the red-brown complex 32c:
205 mg (82%); mp >250 °C (decomp.); 1H NMR (360 MHz, CD2-
Cl2) δ 7.24 (d, J ) 11.3 Hz, 1), 6.82 (s, 1), 6.58 (d, J ) 11.3 Hz,
1), 6.47 (s, 1), 4.62 (s, 5), 3.27-3.22 (m, 1), 3.19-3.15 (m, 1),
2.38, 2.29, 2.26, 2.25 (s, 3 each), 2.20-2.16 (m, 2), 0.86 (s, 3),
0.49 (s, 3); 13C NMR (62.9 MHz, CD2Cl2) δ 139.0, 138.2, 136.9,
134.3, 133.9, 131.8, 131.5, 123.0, 115.5, 112.5, 97.2, 94.7, 88.5,
85.5, 76.8, 32.6, 32.4, 22.5, 21.1, 19.22, 19.20, 19.1, 18.9. Anal.
Calcd for C27H31F6FeP: C, 58.29; H, 5.62. Found: C, 58.18;
H, 5.69.
Rea ction of Br om in e w ith Com p lex 32b To Give
Dibr om id e 34b. One drop of bromine was added to a solution
of iron complex 32b (5 mg) in CD2Cl2 in an NMR tube, and
the spectrum was recorded after 10 min. The peaks for 32b
had then completely disappeared and were replaced by those
of the dibromide 34b: (360 MHz) δ 7.15 (s, 1), 7.04 (s, 1), 6.41
(s, 1), 6.10 (s, 1), 5.57 (d, J ) 5 Hz, 1), 5.23 (d, J ) 5 Hz, 1),
4.76 (s, 5), 3.14-2.64 (m, 4), 2.68 (s, 3), 2.26 (s, 3), 1.01 (s, 3),
0.92 (s, 3).
Rea ction of Br om in e w ith Com p lex 32c To Give
Tr ibr om id e 35c. Bromine (∼500 mg, 3 mmol) in dry dichlo-
romethane (1 mL) was added to iron complex 32c (167 mg,
0.3 mmol) in dry dichloromethane (2 mL), and the mixture
was kept in the dark for 2 days. The product was collected by
filtration and was dried under vacuum to yield 201 mg (94%)
1
of 35c as a red solid: mp >200 °C dec; H NMR (360 MHz,
CD3CN) δ 6.45 (br s, 1), 6.39 (s, 1), 6.24 (br s, 1), 4.75 (s, 5),
3.50-3.40 (m, 1), 3.31-3.22 (m, 1), 2.65-2.54 (m, 2), 2.66, 2.64,
2.60, 2.43 (s, 3 each), 1.28 (s, 3), 0.74 (s, 3); 13C NMR (90.6
MHz, CD3NO2) δ 143.3, 140.2, 137.2, 136.9, 130.4, 127.9, 115.1,
110.2, 101.8, 100.9, 93.1, 86.0, 79.4, 54.2, 51.1, 32.3, 30.3, 22.1,
21.9, 21.4, 21.3, 21.2, 19.8. Anal. Calcd for C27H30Br3F6FeP:
C, 40.79; H, 3.80. Found: C, 40.84; H, 4.00.
Rea ction of br om in e w ith com p lex 32a was carried out
in an NMR tube as for 32b above.
The internal methyl proton and carbon signals for 32a at δ
0.92, 0.67 and 22.3, 20.9 were replaced after 1 h by peaks at
δ 1.18, 1.06, 1.00, 0.97 and 21.0, 20.9, 20.8, 20.7 corresponding
to the two peaks for each of 34a and 35a in about equal
amounts.
Th e Cyclop en ta d ien yl Ir on Com p lexes 32a -c: Gen -
er a l P r oced u r e. A solution of the p-chlorotoluene iron
complex47 33 (176 mg, 0.45 mmol) and the monoene (0.45
mmol) in N2-purged dichloromethane (35 mL) was irradiated
using a 150 W tungsten garden flood lamp for 1.5 h, when a
further portion of iron complex 33 (176 mg, 0.45 mmol) was
added and irradiation continued for a further 2 h. The solvent
was then evaporated, and the residue was preabsorbed and
chromatographed over alumina using dichloromethane/PE/
methanol (55:40:5) as eluant to give the product.
Monoene 7a (106 mg) yielded the brown complex 32a : 176
mg (78%); mp >265 °C dec; 1H NMR (360 MHz, CD2Cl2) δ
7.15-7.12 (m, 3), 7.13 (d, J ) ∼11 Hz, 1), 7.04 (t, J ) 4.8 Hz,
1, H-13), 6.66 (t, J ) 6.1 Hz, 1, H-5), 6.53 (d, J ) 11.1 Hz, 1),
5.96 (d, J ) 6.1 Hz, 1), 5.69 (d, J ) 6.1 Hz, 1), 4.77 (s, 5), 3.11,
3.00, 2.58, 2.42 (see Table 2, H-9,10), 0.92 (s, 3), 0.67 (s, 3);
13C NMR (62.9 MHz, CD2Cl2) δ 139.7, 137.6, 134.7, 130.2,
128.1, 125.4, 124.8, 117.7, 115.2, 89.3, 85.9, 81.6, 76.7, 38.4,
37.4, 22.3, 20.9. Anal. Calcd for C23H23F6FeP: C, 55.22; H, 4.63.
Found: C, 55.03; H, 4.64.
Rea ction of th e Tr ibr om id e 35c w ith Ba se. Sodium
amide (100 mg, 2.6 mmol) was added to a suspension of
tribromide 35c (215 mg, 0.3 mmol) in dry THF (50 mL), and
the mixture was refluxed for 30 min. After cooling, dichlo-
romethane and water were added, and the organic layer was
washed, dried, and evaporated. The residue was preabsorbed
and filtered through a column of SiGel using PE/dichlo-
romethane (4:1) as eluant and gave about 25 mg of white solid
identified as the bromomonoene 36: 1H NMR (360 MHz) δ 6.69
(s, 1), 6.70 (AB, J ) 13.9 Hz, 1), 6.67 (AB, 1), 3.19-3.05 (m, 2,
H
eq), 2.48, 2.40, 2.27, 2.21 (s, 3 each), 2.31-2.12 (m, 2, Hax),
0.83, 0.67 (s, 3 each); EI MS m/z 370, 368 (M+, Br1).
The same debromination occurred with NaOMe/MeOH,
KOBu-t/HOBu-t/THF and DBU.
Attem p ted F or m yla tion of Com p lex 32a . TiCl4 (0.05 mL,
0.45 mmol) was added to a solution of complex 32a (100 mg,
0.2 mmol) and Cl2CHOCH3 (35 mg, 0.3 mmol) in dry dichlo-
romethane (10 mL) at 0 °C and was stirred for 3 h without
further cooling. The mixture was poured on to ice-water, and
more dichloromethane was added. The organic layer was
washed, dried, and evaporated, and the residue was chromato-
graphed over SiGel using PE/CH2Cl2/MeOH (4:5:1) as eluant
and gave first unchanged starting material, 32 mg, and then
2 mg of aldehyde 37: 1H NMR (360 MHz, CD3CN), δ 9.88 (s,
1), 7.69 (s, 1), 7.59 (s, 1), 7.19 (d, J ) 11.0 Hz, 1), 6.69 (d, J )
11.0 Hz, 1), 6.64 (t, J ) 6.2 Hz, 1), 6.00 (d, J ) 6.2 Hz, 1), 5.75
An X-ray structure was obtained on a crystal 0.94 × 0.42 ×
0.36 mm in size, which was found to be monoclinic, space group
P21/c, with a ) 11.636 Å, b ) 12.524 Å, c ) 14.681 Å, and R )
γ ) 90° and â ) 90.69°. The structure was refined to R )
0.0689. Full details of the X-ray determination can be found
in the Supporting Information.
(46) Mitchell, R. H.; Calder, I.; Huisman, H.; Boekelheide, V.
Tetrahedron 1975, 31, 1109.
(47) Khand, I. U.; Pauson, P. L.; Watts, W. E. J . Chem. Soc. C 1960,
2261.

7152 J . Org. Chem., Vol. 64, No. 19, 1999
Mitchell and Zhang
(d, J ) 6.2 Hz, 1), 4.80 (s, 5), 3.27-3.20 (m, 1, H-10eq) 3.05-
2.98 (m, 1, H-9eq), 2.68-2.60 (m, 1, H-10ax), 2.48-2.40 (m, 1,
H-9ax), 0.87, 0.75 (s, 3 each).
be orthorhombic, space group Pccn, with a ) 13.045 Å, b )
14.952 Å, c ) 7.687 Å, and R ) â ) γ ) 90°. The structure
was refined to R ) 0.0657. Full details of the X-ray determi-
nation can be found in the Supporting Information.
Deu ter a tion of Com p lex 29a . n-BuLi (0.30 mmol in
hexane (0.2 mL)) was added to a solution of complex 29a (78
mg, 0.21 mmol) in dry THF (10 mL) at -78 °C and was stirred
for 1 h. D2O (1 mL) was then added cautiously, and then the
mixture was allowed to warm to 20 °C. The solvent was
evaporated, and the residue was preabsorbed on neutral
alumina and chromatographed under N2 using PE/dichlo-
romethane (3:1) as eluant to yield 68 mg (87%) of 38a -c with
an almost identical 1H NMR spectrum to 29a , but 2D NMR
(55.3 MHz) δ 6.37, 5.43, 5.24 (broad singlets, ∼0.5:4:4); in the
13C NMR spectrum, the peaks at δ 118, 95, 92 and 86.5 now
appeared as assymmetric doublets: CI MS m/z 372, MH+ with
D1.
ter t-Bu tyla ted cyclop h a n e 40. t-BuLi (0.11 mmol in
hexane (0.1 mL)) was added to a solution of the complex 29a
(37 mg, 0.1 mmol) in dry THF (5 mL) at -78 °C. After the
mixture was stirred for 1 h, FeCl3/DMF (195 mg) was added,
followed by ether (50 mL). The organic layer was washed,
dried, and evaporated, and the residue was chromatographed
on SiGel using PE as eluant to give the monoene 40, 18 mg
(62%), as an oil: 1H NMR (360 MHz) δ 7.17-6.88 (m, 5), 6.61
(s, 2), 2.95-2.87 (m, 2), 2.53-2.39 (m, 2), 1.36 (s, 9), 0.76 and
0.61 (s, 3 each); CI MS m/z 291 (MH+).
P h otoisom er iza tion of th e Mon oen es 7a -c to 42a -c
a n d Th eir Th er m a l Retu r n . A solution of the monoene (1.2
× 10-4 M) in acetonitrile was degassed thoroughly with argon
for 15 min and then placed in a closed UV cell. The cell was
equilibrated at the desired temperature in a thermostated cell
holder in the UV spectrometer and then was irradiated at 254
nm for 1-2 min. The rate of the disappearance of the
dihydrophenanthrene 42a -c (the thermal return reaction) was
measured by continuously monitoring the absorption at ∼324-
328 nm. In each case the rates were measured at five
temperatures. The graphed data can be found in the Support-
ing Information. A summary of the rate constants obtained
and thermodynamic parameters is in Table 4.
Ack n ow led gm en t. We thank the Natural Sciences
and Engineering Research Council of Canada and the
University of Victoria for financial support.
Su p p or tin g In for m a tion Ava ila ble: Full X-ray struc-
tural data for compounds 41, 8a , and 32a and photoisomer-
ization data (UV) of 7a -c to 42a -c and kinetic data for the
thermal return reaction. This material is available free of
charge via the Internet at http://pubs.acs.org.
X-r a y Str u ctu r e of a n ti-9,18-Dim eth yl-2,11-d ith ia [3,3]-
m eta cyclop h a n e (41). An X-ray structure was obtained on
a crystal4 0.74 × 0.18 × 0.17 mm in size, which was found to
J O990721K