close

Вход

Забыли?

вход по аккаунту

?

Cubanes Starting Materials for the Chemistry of the 1990s and the New Century.

код для вставкиСкачать
Cubanes: Starting Materials for the Chemistry of the 1990s
and the New Century
By Philip E. Eaton*
Dedicated to Professor Horst Prinzbach on the occasion of his 60th birthday
The study of non-natural products has led to a broad understanding of bonding and reactivity
in organic chemistry. Many times, compounds thought impossible have been realized in the
course of such studies. Cubane, a landmark in the world of “impossible” compounds, has been
found to have a rich chemistry, full of the unexpected. The recent renaissance of cubane
chemistry, triggered by potential applications of the system to the production of high-energy
fuels and the like, has led to many discoveries including the first methods for systematic
substitution on strained, saturated systems and a new process for the metalation of arenes,
ortho magnesiation. Reactive intermediates with exceptional bonding parameters have been
uncovered and characterized including 1(9)-homocubene, the most twisted olefin; cubene, the
most pyramidalized olefin; cubyl cation, once the “least likely” cation; cubylmethyl radical,
a saturated radical that rearranges on the picosecond timescale; and many other extraordinary
species. There is certainly good reason to believe that future work in the cubane arena will be
at least as productive (probably more so), and that it will help develop a deeper understanding
of chemistry.
1. Introduction
Cubane I is immensely strained; the geometry at each
carbon atom is far from tetrahedral. Originally there was
doubt that the skeleton could even hold together.[’] Only
later was it appreciated that no kinetically viable paths exist
for the thermal rearrangement of cubane. On one hand, orbital symmetry considerations show that the energy of concerted two-bond ring-opening reactions is very high.12as
b1 On
the other, there is little to be gained by breaking just one
bond, as the concomitant change in geometry is small, and
the resulting biradical is still very strained.[*’]
2. Entry into the Cubane System
In 1964, at The University of Chicago, Tom Cole and I
achieved the first synthesis of the cubane system
(Scheme
The Favorskii rearrangements (semibenzilic
mechanism) by which the bishomocubane system is contracted were the critically important steps. This is true to this
day for most cubane syntheses; all but two depend on this
reaction.r41
Even in retrospect, the whole sequence is really rather
good, although the early part of the procedure requires the
[*I
Prof. P. E. Eaton
Department of Chemistry
The University of Chicago
5735 South Ellis Avenue, Chicago, IL 60637 (USA)
Angew. Chrm. I n f . Ed. Engl. 1992, 31, 1421-1436
0 VCH
- LacooH
NaOH/H20
a’
HOOC
2
Scheme 1. The first entry into the cubane system. a) Reflux. NBS = N-bromo
succinimide.
preparation of 4-bromocyclopentenone, a tricky matter. A
few years after the original synthesis was published, N. B.
Chapman et al. in
following up on my work
with Hudson on bromocyclopentadienone ketals,L6] found
that the ethylene ketal of 2-bromocyclopentadienone could
be made easily and undergoes spontaneous dimerization.
From this discovery the cubane synthesis ultimately evolved
into a simple five-step process providing cubane-l ,4-dicarboxylic acid 2 in about 25 ‘3’0 overall yield (Scheme 2).
2.1. Availability of Cubanes
The synthesis of 2 has been scaled up and is conducted in
small pilot plants by Fluorochem in California and EniChem
Synthesis in Milan. Cubane-l,4-dicarboxylic acid is being
Verlagsgesellschafi mhH. W-6940 Weinheim.1992
0570-0833~92~11ll-1421
S 3.50+.25/0
1421
1. NaOH
2. H+/H20
in very nearly quantitative yield. It is beautifully crystalline;
alas, the crystals are rhombohedral, not cubic.
3. Physical Properties
n
Because of its exceptional structure, strain, and symmetry,
cubane is a benchmark in organic chemistry. Cubane has
been studied extensively, and much concerning its properties
has been published.['] A few physical parameters are given in
Table 1.
Table 1. Selected data for cubane.
2
Scheme 2. The best synthetic route known for the cubane system. a) Glycol,
H'; b) reflux.
made in multi-kilogram batches! Nonetheless, as is often the
case with experimental compounds, cubane 1 and its derivatives are still expensive to purchase. Fortunately, any good
student, having gone through the procedure once to become
familiar with its details, can easily make 10 grams of 2 in a
month or so. It is a very satisfying procedure and an excellent
way to introduce students to the delights of distillation and
the thrill of crystallization. It is unfortunate that in today's
academic synthesis laboratories these joys are often missed.
2.2. The Hydrocarbon
The original method for the decarboxylation of cubanedicarboxylic acid 2 to provide the pure hydrocarbon itself proceeded via the tert-butyl p e r e ~ t e r . ~
The
~ ] present method is
much superior. Following from the lovely work of Derek
Barton et al.,[71we now use the radical-induced decomposition of diester 3, which is prepared simply from 2. Cubane,
the hydrocarbon, is thus obtained easily on a 10-gram scale
U
3
1
Scheme 3. The decomposition of 3 provides 1 in good yields. a) Reaction in
THF at reflux; irradiation (tungsten filament).
C-C distance
C-H distance
Color
Toxicity
Stability
Decomposition
Density
Vapor pressure
1.5727 f 0.0019 8, [a]
1.138 f 0.008 A
transparent [b]
nontoxic
inert to light, water, air
> 220 "C [c]
1.29gcm-'
1.1 mm (25 "C)
Boiling point
5 133 "C
Melting point
130-131 'C
Solubility
~ 1 wt%
8 (hexane)
Heat of formation [d] 144 kcalmol-'
( + 159 kcalmol- ')
Strain energy [dl
166 kcalmol-'
(181 kcalmol-')
+
[a] Old value 1.551 A. [b] To below 200 nm. [c] Very slow. [d] Values not in parentheses
from Margrave et al. [12]; values in parentheses from Domalski et al. [13].
The bond lengths given in Table 1 are from a recent electron diffraction study by L. and K. Hedberg et al.[91 and
agree reasonably well with the earlier data of Almenningen
et a1.["] The C-C bond length is a bit longer than that obtained in the original (and crucial) X-ray structure determination by Fleischer in 1964.[1'a1It is interesting that this
distance is not much different than the C-C bond length in
a simple cyclobutane.["bl The values for the heat of formation and strain energy given in parentheses below the results
of Margrave et a1.L'21were derived by Domalski et al. at the
U.S. National Institute of Standards from the heat of comWe hope
bustion of 1,4-di(methoxycarbonyl)c~bane.~~~~
soon to have new data directly from combustion of
cubane.
Molecular mechanics calculations, and semiempirical and
ab initio quantum mechanical treatments of cubane abound.
Robiette et al. have provided an excellent overview.191The
agreement between methods is fair, but to my mind certainly
less than completely reassuring. Whatever one's opinions
about the calcukated properties of cubane, the bottom line
experimentally is that cubane is remarkably stable. Cubane
melts just above 130"C; measurable decomposition occurs
only above 200 "C. Martin et al. have shown that the energy
of activation for the therrnolysis of cubane at 230-260 "C
Philip E. Eaton was born in New York in 1936. He received his undergraduate degree from
Princeton University in 1957, then did his Ph. D . work with Peter Yates at Harvard University,
graduating in 1961. From 1960 to 1962 he was an assistent professor of chemistry at the University of California at Berkeley. He then accepted a position at The University of Chicago, where
he is a full professor. Eaton has received the Rohm and Haas Award for Research, an
Alfred P. Sloan Foundation Fellowship, and a Senior Scientist Award from the Alexander von
Humboldt Foundation. His research interests include the synthesis of important non-natural
products and their applications in materials science.
1422
Angew. Chem. Int. Ed. Engl. 1!292,31,1421-1436
is 43.1 1 .O kcalmol-’, an extraordinarily high value.[2b1
Cubane is entirely stable to light, air, water, and most common reagents.
4. The Cubane Renaissance
Although many physical properties of cubane have been
measured, until a few years ago cubane was considered just
a laboratory curiosity of interest only to academics. That
changed in the early 1980s when Gilbert of the U.S. Army
Armament and Development Command (now ARDEC)
pointed out that cubane’s very high heat of formation[”]
and its exceptionally high density (1.29 gcm-3)[11a1-higher
than almost all other hydro~arbons[~~~--could
make certain
cubane derivatives important explosives. His colleagues
Sandus and Alster at ARDEC provided theoretical support
and estimated that octanitrocubane 4 might be 20-25%
more effective than HMX (octogen), the present-day standard.[l5]It is sufficient to comment here that detonation of
cursor for polyurethanes. Such materials may find application as high-energy binders for propellants. Cubylamines
(and more particularly cubylmethyl amines) are of interest as
antiviral agents given the relation of their size and shape to
the important drugs Symmetrel LC and Rimantadine y. (DuPont). 1,4-Dinitrocubane, the final product in Scheme 4, is
perfectly easy to handle. It is less shock-sensitive than TNT
2
Scheme 4. Functional group transformations on the cubane framework
a) Warming; b) neutralization.
octanitrocubane would lead to an awesome amount of very
hot gas [Eq. (a)].
C,(NO,),
~
~
t
8 CO,
+ 4 N,
(a)
Miller at the U.S. Office of Naval Research realized quickly that some substituted cubanes could be very useful propellants for the space and “Star Wars” programs. The synthesis
of 4 and related compounds are formidable tasks. The challenge made us aware that we did not know how to prepare
substituted cubanes beyond a very few simple examples.[41
At the time we returned to cubane chemistry in 1983 there
was little methodology for taking the most readily available
cubane compound, the 1,4-dicarboxylic acid 2, and doing
anything besides manipulating the carboxyl groups.
5. Functional Group Transformations
As it turned out, the carboxylic acid groups of 2, like most
other functional groups on cubane, are exceedingly wellbehaved. Most typical functional group transformations can
be applied successfully. This can be appreciated in the preparations of 1,4-dinitrocubane worked out at Chicago by
Shankar et al. and Wicks et al. (Scheme 4).[16].Here, modern
versions of classical methodology are used. These conversions are as easy to do on the cubane system as on any
standard carboxylic acid. Note, just as one example of the
“convenient” behavior of most cubane derivatives, that the
intermediate cubyl-I ,I-diisocyanate in Scheme 4 is a stable,
crystalline material (m.p. 112-114 “C) and an obvious preAngew. C‘hem. Inr. Ed. Engl. 1992, 31, 1421-1436
and remarkably stable thermally (decomp. 260 “C). [This is
not true of some other cubanes; cubyl azides, diazoketones,
and diazomethanes are shock sensitive and dangerous.
Caution is advised in handling any cubane, since all are highenergy compounds. Safety shields should always be used,
and particular care should be taken in the workup of crude
mixtures that might contain acidic or heavy-metal contaminan ts .]
6. Substitution on the Cubane Framework
Functional group transformations are useful for preparing many mono- and disubstituted cubanes, as the corresponding carboxylic acids are readily available.[31But this
does not suffice for the preparation of more highly substituted cubanes. As Reddy et al. have shown, substitution of the
cubane framework by way of the cubyl radical is fairly easy.
However, as is usually the case with radical reactions, mixtures of products are obtained. Light-induced iodination of
cubane with tert-butyl hypoiodite gives mono-, di-, triiodocubanes, etc.[’’] The challenge is to achieve systematic and
controlled substitution onto the cubane frame. This requires
a real understanding of the structure of cubane.
6.1. The Nature of the C-H Bond in Cubane
The geometry of the cubane system dictates substantial
rehybridization of the component tetravalent carbon atoms
away from sp3 hybridization. The carbon-carbon bonds become p-rich, and the exocyclic carbon orbital used for the
C-H bond becomes s-rich. There is about 31 % s character in
this orbital as calculated from the observed %-‘H coupling
‘I
Thus, the hydrogen atoms of
constant of 3 55 HZ.‘~,
1423
cubane are expected to be more acidic than those of strainfree, saturated hydrocarbons. This has been demonstrated
decisively.[191The kinetic acidity of cubane is about 63 000
times that of cyclohexane. Nonetheless, this is far too low to
be useful for direct synthetic applications.
6.2. Amide Activation for ortho Metalation
Our idea was to use neighboring group activation to assist
metalation of cubane. It was already well known from the
work of Meyers, Beak, Snieckus, and others that an amide
substituent on an arene aids lithiation in the ortho position
by activation of the arene hydrogen atom.[*’] There are certain similarities between cubanes and arenes: both have C-H
bonds with enhanced s character, and in both the adjacent
(ortho) substituents are forced to be coplanar (Scheme 5).
Thus, we speculated that it might be possible to lithiate
cubanamides.
presence of a huge excess of base is not an easy thing to work
with.
6.3. Transmetalation
In a key step forward, Castaldi showed transmetalation to
be the golden link in the synthesis of substituted cubanes.121]
When lithiation of a cubanamide is performed in the presence of mercury salts, the lithium compound first formed is
quickly mercuriated, and the very polar C-Li bond is replaced by the more nearly covalent (and relatively inert)
C-Hg bond. In this way, the starting material 5 can be completely converted into a metalated cubane (Scheme 6). The
5
H
/
‘L33%S
-
Hg-IScheme 6. Metalation/transmetalation: a method for the complete metalation
of the activated cubane framework.
31%~
Scheme 5. A comparison ofimportant features of benzamide, cubanamide, and
their orrho-lithiated derivatives. The similarity between arenes and cubanes is
reflected in the similar CCH angle and amount of s character in the C-H bond.
The seminal work was done in my laboratory by Graziano
Castaldi.[”’ He showed that a small equilibrium concentration of the ortho-lithiated derivative 6 is indeed formed when
the diisopropylamide 5 of cubanecarboxylic acid is treated
with a large excess of a strong, non-nucleophilic base like
lithium tetramethylpiperidide (LiTMP). (Alkyllithium
reagents are not suitable because they attack the amide
group.) When the reaction mixture was quenched with
CH,OD the recovered cubanamide contained approximately 3 % deuterium in the 2 position (ortho); dideuterated material could not be detected
C(O)N(i Pr),
C(O)N(i Pr),
LiTMP
H ‘
-
5
CH3OD
-QLi
6
This result was very satisfying to us as physical organic
chemists since it was logically derived and demonstrated
that activated saturated systems can be metalated.[22,231
Nonetheless, this observation alone was not particularly useful in synthesis. A small concentration of a carbanion in the
1424
amide group is important in stabilizing the intermediate
lithiated cubane, but not the mercuriated compound. Once
lithium is replaced by mercury, the amide group is again able
to assist removal of another ortho-hydrogen atom, and the
process can continue.
Metalation/transmetalation is the basis of a complete synthetic methodology for the preparation of a great variety of
substituted cubanes. Mercuriated cubanes are useful for the
formation of halocubanes, as the C-Hg bond is readily
cleaved by halogens such as iodine. Other metalated cubanes
are more reactive and thus more useful. Higuchi etal.
showed that zinc, silicon, and tin derivatives of cubane could
be made in a similar way (Scheme 7) and are of great synthetic
M = Si,Sn
Scheme 7. Preparation of silyl- and stannylcubanes by metalation/transmetalation.
Angew. Chem. Int. Ed. Engl. 1992,31, 1421 -1436
6.4. Reverse Transmetalation
traordinarily good; in the example shown in Scheme 9 yields
of 90-95 % are easy to achieve in both small- and large-scale
Cunkle, a postdoc in my group, worked out a reverse
transmetalation methodology for converting mercuriated
cubane compounds into more reactive metallocubanes.[25”]
This extraordinary (if roundabout) process allows the quantitative preparation of dilithiated cubanes (Scheme 8), materials not available by direct lithiation of cubanamides. The
method has also found use in the synthesis of dimetalated
arene~.[’~~I
-
LiTMP-
( i Pr),NC(O)
H&J;C)”Pr)t
(i Pr),NC(O)
transmetalation
I
LiTMP/HgClp
t
6.6. The Amide Activating Group
The diisopropyl amide activating group used in these various reactions was chosen because it is quite inert to the
amide bases employed for ortho metalation. As always, there
is a price to be paid for the luxury of judicious choice: such
amides cannot be hydrolyzed to the corresponding carboxylic acid, the functional group most amenable to further
transformations. For a time this problem was the “Achilles
heel” of the reaction sequence. We did find finally that borane reduction of the amide followed by oxidation of the
amine so produced with dimethyldioxirane (Scheme 10)[27a]
or (on larger scale) with potassium pennanganate provided
a successful
qIr HmH
c--(i Pr),NC(O)
Hg-I-
HOOC
p LiAIH4
r ) 2
(i Pr)zNC(0)
HzN(iPr)2
(iPr)dCH2
H20H
,COOH
H O O C T
COOH
Scheme 8. Metalation/transmetalation/reversetransmetalation: a roundabout
route to not easily accessible compounds.
6.5. Cubyl Grignard Reagents
Bashir-Hashemi at Geo-Centers introduced transmetalation with magnesium salts and thereby provided easy access
to cubyl mono- and bis-Grignard reagents.[261This works
exceedingly well. The conversion proceeds by stepwise lithiation/bromomagnesiation. The second metalation of the
Scheme 10. A novel way to “hydrolyze” a recalcitrant amide group
Bottaro, Schmitt, etal. introduced later the use of the
tert-butyl ethyl amide as an activating group for ortho metalation that could be completely hydrolyzed in strong acid.g2s]
Subsequently, we found that just the tert-butyl group alone
can be removed in mild acid, and then the resulting amide (in
our case a methyl amide) can be taken to the free carboxylic
acid with either acid- or, better, base-catalyzed hydrolysis.
6.7. ortho Magnesiation
In the course of extending our work to the synthesis of
cubanes with different substitution patterns, we found that
treatment of the 2,4-dicyanocubanamide 7 with LiTMP
gives two isomeric tetrasubstituted cubane carboxylic acids
(8 and 9) after c a r b o ~ y l a t i o nThe
. ~ ~ electron-withdrawing
~~
COOH
Scheme 9. ortho-Magnesiation by metalation/transmetalation: a useful method.
(O)N(iPr)z
8
,,
LiTMP
+
____)
cubane framework occurs at a position remote from the
first-quite understandable considering the polarities of the
intermediates. Further metalation does not take place. In
most cases, the overall conversion of these reactions is exAngew. Chem. Int. Ed. Engl. 1992, Sf, 1421-1436
NCJ@N
7
2. c
q
3. H3O+
NC
1425
effects of the cyano groups along with the amide stabilizes
both intermediate lithiated cubanes so well that only a small
excess of LiTMP is needed to achieve fairly complete deprotonation even at -78°C. In the initial lithiation step, the
activation of a cubane C-H bond by two flanking cyano
groups rivals the ortho activation by the one amide group.
This results in competitive lithiations and a mixture of products. In attempts to achieve greater selectivity in this reaction, we examined the effect of added MgBr,. When the ratio
of LiTMP to MgBr, was 1.5:1, the product ratio was improved to 9 : l favoring carboxylation ortho to the amide
function. This suggests an important, previously unknown
reaction: direct magnesiation by TMPMgBr (a Hauser
base)[301and/or by (TMP),Mg. Lee and Xiong in my laboratory have expanded this idea into a powerful new process
dubbed ortho magnesiation, and we have found that magnesium amide bases are exceedingly useful in synthe~is.'~'.
321
These bases can be made very simply. The reaction of
2,2,6,6-tetramethylpiperidine (TMPH) in refluxing T H F
with one equivalent of ethylmagnesium bromide or with
0.5 equivalents of dibutylmagnesium gives, respectively,
clear solutions of TMPMgBr and (TMP),Mg. The reactions
are complete within a few hours. Solutions of the corresponding diisopropylamine derivatives can be obtained simi l a r l ~ . [ ~These
''
magnesium amide bases are stable in refluxing T H F for hours. In this way they differ markedly from the
corresponding lithium amides, which rapidly decompose at
this temperature; indeed, LiTMP is unstable above 0 "C in
T H E Since the magnesium amide bases can be used in refluxing THF, reactions with materials of otherwise low solubility or low reactivity are possible.
When dicyanocubanamide 7 is treated with TMPMgBr or
(TMP),Mg the only product after carboxylation (90%
yield!) is 8 in which the carboxyl group is next to the amide
activating group.1291This reaction proceeds well even at
- 78 "C. The inductive effect of the cyano group clearly enhances the reaction, but the cyano group itself is not an ortho
director for magnesiation.
C(O)N(i
1. TMPMgBr/
FnP)flg
2. co2
3. K O +
NCN
C
N
7
6" 0 ::i 6
COOMgN(i
Mg[N(jPr)&
gN(iPr)z
____t
/
/
(iPr)2NMg
COOH
HOOC
COOMgN(i
'
COOH
10
and further extensions of the ortho magnesiation process are
discussed in more detail elsewhere.[31.3 3 - 3 5 1
7. Synthesis of Cubanes with Different
Substitution Patterns
Although activation by a single amide group suffices for
the metalation of cubanes, the presence of additional electron-withdrawing groups enhances reactivity enormously.
4-Cyanocubanamide is much more reactive than the unsubstituted amide. Magnesiation occurs quickly and exclusively
ortho to the amide. As Puranik, a postdoc in my group, has
shown for certain cases, namely when the electrophile can
coexist with the base, it is possible to substitute all three
positions ortho to the amide in a wonderfully simple, one-pot
reaction proceeding by sequential metalations. Thus, for example, 4-cyanocubanamide can be converted directly into
the tri(tert-butylcarbonyl) derivative 11.[361
C(0)t Bu
tBuC(0)
m C ( O ) N C H a ( t Bu)
NC
C(O)NCHB(f Bu)
+%)cH3
TMPMgBr
11
As the Baeyer-Villiger oxidation of tert-butyl cubyl ketones is completely specific-only the tert-butyl group migrates-triketone ll can be converted easily to the polycarboxylated cubane 12.1361
g N Y O ) N ( i Pr),
*
tBWO)
C(0) t Bu
C(O)NCHs(t Bu)
NC
8
NC
OOt Bu
11
The uses of ortho magnesiation extend far beyond the
cubane system; for example, the method can be applied productively to aromatic systems. "By-products" such as this
have arisen often from our work on the cubane system and
as in this case have brought new, important insights. Since
the original work by Gilman on ortho lithiation of substituted arenes, it has always been assumed that the metalation
depends crucially, if not exactly explicably, on the properties
of lithium. Now we know that direct ortho magnesiation can
be accomplished as well. Such processes complement ortho
lithiation and sometimes also offer distinct advantages, as
shown in the new synthesis of pyromellitic acid (10). This,
1426
HOOC
12
We have used processes like these to produce many different cubanes. It is fair to say that almost any combination or
permutation of functional groups and positions on the
cubane nucleus can be achieved now by fairly simple chemi ~ t r y . ' In
~ ~Scheme
]
11, for example, are six tetrasubstituted
Angcw. Clzcm. h i . Ed. EngL 1992, 31, 1421-1436
GG-1
(=TGF]
(1,2,3,6-)
WP
. .. .
1,2,3,8-)
NC
HOOC
C(O)N(i Pr),
NC
(7GT-I
(W)
9. Cubanes as Precursors to Reactive
Intermediates
Having solved the problems associated with the synthesis
of substituted cubanes, we can now start with the cubane
system, itself an extraordinary structure, and move outward
in our continuing exploration of the limits of bonding in
organic compounds.
ooc
9.1. Cubyl Cation
C(O)OCH,
Scheme 1 1 . Tetrasubstituted cubanes with different substitution patterns.
cubanes we have made, each with a different substitution
pattern.
We have even synthesized cubanes like 13 with fused heterocyclic
Quite different from expectation these
“heteropropellano cubanes” are remarkably stable. The ni-
One of the biggest surprises we have had in our work was
the discovery that cubyl cation is much more accessible than
expected.[391Everything about cubyl cation seems unfavorable: 1) The geometry about the positively charged carbon
atom is very far from flat; 2) the exocyclic orbitals in cubane
are s-rich; 3) hyperconjugative stabilization would require
high-energy, cubene-type structures. Ab initio calulations
(6-31G*, without electron correlation) place cubyl cation
about 20 kcalmol-’ higher in energy than tert-butyl cation
and about 5 kcalmol-‘ higher in energy than 1-norbornyl
cation.[401Nevertheless, we came across numerous reactions
(Scheme 12) that might involve the intermediacy of the cubyl
13
trogen atoms can be alkylated or nitrated without interference from ring-opening reactions. Interestingly, the C-C
bond shared by the urea ring and the cubane system is substantially shorter than the equivalent bond in cubanes or
cyclic ureas. The origin of this effect is not known.[381
8. Cubane as a Building Block for
Pharmaceuticals
Because the cubane frame is rigid, substituents have precise spatial relationships to one another. The distance across
the cube (the body diagonal) is almost the same as that
between the para positions of the benzene ring. On cubane,
one can add substituents in “the benzene plane”, as well as
above and below it, so to speak. This offers fascinating possibilities for the synthesis of new pharmaceuticals. A number
of cubane derivatives have already been obtained which
show interesting activity in anti-AIDS and anti-tumor
screens. Admittedly, the activity/toxicity balance is not yet
satisfactory. We do know, however, that the cubane system
is not inherently toxic; most cubanes are biologically innocuous. The search for pharmaceutically significant cubanes has
just begun. At this point, cubane should be viewed as a
biologically stable, lipophilic platform on which the chemist
can install a wide choice of substitutents in a variety of welldefined spacial relationships. Developments in drug design
programs should allow the judicious choice of substituents.
AnKen. Chem / n t . Ed. Engl. 1992, 31. 1421 -1436
W O C H 3
Scheme 12. Substitution reactions of cubane derivatives that probably proceed
via the cubyl cation.
9.1.1. Solvolysis of Cubyl Triifrate
Since some of the substitution reactions in Scheme 12 do
not have clearly defined mechanisms, and there are possibilities for front-side attack, “hot intermediates”, etc., we
turned to an examination of simple solvolysis reactions.[39a1
As it turned out, the solvolysis of cubyl triflate 14 in pure dry
methanol is remarkably facile. Its half-life at 70 “C is only
14
15
15 minutes. The reaction is very clean. No rearrangement is
observed, and the only product is cubyl methyl ether 15.
Labeling studies have established that the methoxy group is
1427
attached to the same carbon as the original triflate group.
There is no hydrogen scrambling nor skeletal rearrangement.
carbon atoms (C3). Population analysis reveals that the
y-CH (C4) and IX-CH(C2) groups bear considerable positive
charge (0.16), more than the P-CH groups.
9.1.2. A Comparison of Experimental and Calculated
9.1.4. Substituent Effects on the Rate of Solvolysis of
Cubyl Tripate
Solvolysis Rates
Cubyl triflate is very much more reactive than 1 -norbornyl
triflate which is not solvolyzed noticeably by methanol at
I 0 "C even after 250 hours. In hexafluoroisopropanol at
60 "C norbornyl triflate is only 90 % solvolyzed in 85 hours,
whereas cubyl triflate is converted completely in this solvent
to cubyl hexafluoroisopropyl ether within 5 minutes at room
temperature.
The logarithm of the S,1 solvolysis rate is, in a first approximation, inversely related to the difference in the strain
energy of the non-ionized starting material and the intermediate cation. Schleyer et al.[421have shown that this energy
difference can be estimated successfully by careful application of molecular mechanics calculations. In Table 2 published rate data for solvolysis have been adjusted (in so far as
possible)[431to a common set of conditions, and are compared to the values predicted by molecular mechanics. The
Table 2. Comparison of the experimentally determined and the calculated relative rates of solvolysis of some tertiary tosylates in acetic acid at 70 "C.
Tosylate
terf-butyl
I-adamantyl
l-bicyclo[2.2.2]ocry1
Exp.
1
lo-'
Calc.
1
Tosylate
Exp.
Calc
1-norbornyl
cubyl
10- '0
< 10
lo-"'
~
lo-'
agreement is within one or two orders of magnitudeastonishing considering the substantial corrections. Cubyl
tosylate is the exception; here the theoretical and experimental rates of hydrolysis differ by more than 15 orders of magnitude!
9.1.3. The Bondng in Cubyl Cation
Why is cubyl cation so much more easily formed than
expected?[44] Put another way, what differentiates the
cubane derivative from the other compounds in Table 2? As
discussed earlier in Section 6, the geometric constraints inherent in the cubane skeleton force significant rehybridization in the bonding orbitals of carbon away from sp3. The
C-C bonds in cubane are p-rich in comparison to those in
the other systems listed in Table 2. The electronic structure of
the bonding system in cubane is quite different from that of
the others; molecular mechanics does not take this into account.[451The matter has been treated quantitatively by Borden and Hrovat with ab initio 6-31G* calculations including
electron correlation.[401This has led to the conclusion that
the positive charge is delocalized by interaction with the
strained C-C bonds in cubane; namely, there is positive
bond order between the p orbital at the positively charged
carbon atom and the p orbital aligned with it at each of the
1428
Considering the calculated charge distribution in cubyl
cation, it is not surprising that electron-withdrawing substituents at position 4 (body diagonal, para) suppress solvolysis effectively (Table 3). We were unable to detect any signifTable 3. Comparison of the half-lives rliz for the hydrolysis of a number of
4-substituted cubyl triflates.
4-X-1-Triflate T ["C]
t1,2
4-X-1-Triflate T [ "C]
fll2
X=H
CH,O
C1
CF,SO,
< 5 min [a]
>I33 h [a]
>20 d [a]
[b]
H
CH,
(CH3),Si
(CH313Sn
67 h [c]
200h [c]
37 min [c]
[dl
20
60
60
100
25
25
25
25
[a] In (CF,),CHOH. [b] Reaction time in (CF,),CHOH was too long to measure. [c] In CH,OH. [d] Reaction time in CH,OH was too short to measure.
icant solvolysis of the bistriflate of cubane-1,4-diol after it
was heated in methanol at 85 "C for 5 days, in aqueous 80 %
ethanol at 100 "C for 5 days, or in hexafluoroisopropanol at
100°C for 6 days.[461On the other hand, as Borden et al.
insightfully pointed
n-donor substituents cannot stabilize cubyl cation effectively, since the C-to-substituent
bond at C4 is along the axis of the cylindrically symmetric
LUMO of the
Thus, substituents such as methyl
and methoxy normally associated with the hyperconjugative
and resonance stabilization of cations fail to do so in cubyl
cation. Indeed, as can be seen in Table 3, only their destabilizing inductive effects are mirrored in the observed rates.
The symmetry restriction limiting the stabilizing effects of
ndonor substituents of course does not apply to r~ donors.
Jian Ping Zhou in my laboratory has found that silyl and tin
substituents at C4 markedly enhance the rate of solvolysis of
cubyl triflate~.[~']
Indeed, solvolysis of 4-trimethylstannylcubyl triflate in methanol is so fast (about 3000 times faster
than cubyl triflate) that its rate could only be obtained
by extrapolation from the reaction of the corresponding
mesylate.
9.2. Dehydrocubanes
Cubane is a saturated hydrocarbon of extraordinary geometry and strain. Although it was once dismissed as impossible, it has been shown to be stable. Similarly, cubyl cation
has been found to be far more readily accessible than once
imagined. With these lessons in mind, it is prudent not to
dismiss other uncommon notions out of hand.
9.2.1. 1,2-Dehydrocubane (Cubene)
1,2-Dehydrocubane, cubene 16, is the most extreme ex,
the geample of a pyramidalized ~ l e f i n . ["1~ ~Obviously,
Angew. Chem. Int. Ed. Engl. 1992, 31, 1421-1436
ometry about the vinyl carbon atom can be nowhere near
planar. Hrovat and Borden have calculated a pyramidalization angle of 84" for 16.150b1
thracene, the 1:1 Diels-Alder adduct 17 was formed in 64 %
isolated yield. Cubane-on-a-pedestal!
t Buli
Using the newly developed methods for the synthesis of
substituted cubanes discussed earlier, Maggini, a postdoc in
my group, prepared 1,2-diiodo~ubane.[~~~
He found that the
chemistry of this vicinal dihalide could be rationalized
most efficiently by invoking the intermediacy of cubene
(Scheme 13).[511
Critical to the reasoning behind this hypoth-
I'
16
'I
I
I
1.fBuli
2.CH3OH
1
%-
CH30H
Scheme 13. The reaction of
1.2-diiodocubanewith
Engf. 1992,31, 1421-1436
The extraordinary chemistry of 1,2-diiodocubane induced
us to examine the 1,Cdiiodo compound,1521a material readily available from cubane-I,4-dicarboxylic acid.[531In my
laboratory, Tsanaktsidis showed that the reactions of 1,4diiodocubane with alkyllithium reagents (Scheme 14) give
rise to products related to those obtained from the reactions
of 1,2-diiodocubane, that is, materials resulting from the
nucleophilic additions of lithium compounds to some very
reactive
Our labeling experiments provide good evidence that the
products arise from a symmetrical intermediate. There are a
number of possible structures, for example 18-21, but some
of these can be eliminated from serious consideration. 1,4Dilithiocubane (18) can be prepared if the exchange reaction
between 1,4-diiodocubane and tert-butyllithium is conducted at - 100 "C; 18 behaves as would be expected, quite differently from the intermediate generated in the reaction performed at - 70 "C. It is not likely that Grob fragmentation
of the diiodide gives 19, surely a reactive olefin, since the
addition of suitable trapping dienes like furan and 9,10diphenylisobenzofuran does not change the observed course
of the reaction. Borden et al.[551and Michl et al.[561have
calculated independently that structure 20 with a body-diagonal bond is unlikely given its exceedingly high strain and
fBuLi.
esis is the fact that reaction of monoiodocubane with tertbutyllithium gives essentially only cubyllithium ; no significant amount of tert-butylcubane is formed. The formation
of tert-butylcubane and cubylcubanes in the reactions of
1,2-diiodocubane presumably reflects the susceptibility of
cubene to nucleophilic attack. Certainly, nucleophilic addition to highly strained olefins is well precedented in the lovely work of Szeimies et al. on bridged bicyclob~tenes.[~~]
Cubene is an immensely strained, relatively "open", and
sterically unhindered olefin. As such, it can be expected to be
very reactive in Diels-Alder additions. Indeed, when the
deiodination of 1,2-diiodocubane was conducted at room
temperature with tert-butyllithium in benzene in the
presence of 11,i 2-dimethylene-9,1O-dihydro-9,10-ethanoanAngew. Chrm. Ini. Ed.
9.2.2. 1,4-Dehydrocubane (1,4-Cubanediyl)
II
P
\
19
20
21
Scheme 14. Possible intermediates 18-21 in the reaction of 1,4-diiodocubane
with rBuLi.
1429
orbital phase mismatch.[571Ab initio calculations by both
groups indicate that the actual intermediate is probably 1,4cubanediyl (21).
The singlet state 22 of 1,4-cubanediyl is calculated to be
> 10 kcalniol-’ more stable than the triplet state, a result of
substantial through-bond interaction. Indeed, 1,4-cubane-
I
X= I, Br, CI
Ph
Phli
Ph
Scheme 16. The reactions of 4-halo-1-iodocubanes with phenyllithium
diyl does not behave like a diradical. It does not react with
cyclohexene; neither hydrogen abstraction from, nor addition to the olefin is observed. Michl et al. have obtained
spectroscopic evidence for cubanediyl by matrix isolation
(argon, 12 K) of the products from the gas-phase metalinduced dehalogenation of 1,4-dihalo~ubanes.[~~~
t Buli
l-7
NLi
Ph
10. Cubylcubanes and Oligocubanes
Cubene (1,2-dehydrocubane) and 1,4-cubanediyl (1,4-dehydrocubane) are enormously strained compounds which
cannot be isolated under standard conditions. They both
undergo nucleophilic additions very readily, and this has
provided us access to the cubylcubane series (see Schemes 13
and 14). Gilardi[581has shown by X-ray structure analysis
that the central cubylcubane bond is exceedingly short
(1.458 A, Scheme 15) as might be expected, since the exocyclic orbitals of cubane are s-rich and close to the nucleus.[59]The central bond in cubylcubane has about the same
t
flL
Ph
Scheme 17. The key steps in the synthesis of arylated [n]cubylcubanes.
The [n]cubylcubanes have rigid, rod-shaped skeletons and
allow the preparation of sets of compounds with interacting
groups separated in space by precise distances. Each cubane
unit adds 4.15 8, to the length of the molecule. Pramod developed smooth methods to place different arene substituents at opposite ends of cubylcubanes. Thus, for example, he could trap the product from Scheme 17, the
lithiocubane with n = 2, with phenanthrene-9,lO-epoxide
and then dehydrate the resulting intermediate to give 23.r621
Scheme 15. The molecular structure of cubylcubane. The C-C bond lengths
were determined by X-ray structure analysis [%I.
length as the C-C single bond made from sp2 orbitals in
1,3-butadiene. Note that the central bond in adamantyladamantane, constructed from sp3 orbitals, is 1.578 8,
10.1. Arylated Cubylcubanes
T~anaktsidis‘~~]
found that 4-haloiodocubane reacts with
phenyllithium to give, ultimately, I-iodo-4-phenylcubane.
He demonstrated that the reactions proceed by way of the
1,Cdiyl (Scheme 16).[491
K. Pramod extended and modified the sequence in a way
that permits the preparation of a host of [n]cubylcubanes
(Scheme 17).[61,621
1430
23
The group of Gerhard Closs (deceased May 24, 1992), my
late colleague at The University of Chicago, is now measuring the rates of electron transfer between such substituents as
a function of the distance separating them.
Angrw. Chem. I n t . Ed. Engl. 1992, 31, 1421-1436
10.2. Alkylated Cubylcubanes
The preparation of alkyl-substituted cubylcubanes in
good yield by the addition of alkyllithium reagents to
cubanediyls is more difficult than the synthesis of the aryl
analogues. Nucleophilic addition reactions most probably
proceed by way of single-electron transfers, and a good
match of the orbital energies is required for this to be effective. Aryl- and cubyllithium add easily to 1,4-cubanediyl, but
alkyllithium reagents do not. Yusheng Xiong in my laboratory found a way around this limitation by making intelligent
use of Grignard intermediates as shown in Scheme 18.[631
ways. The successful esterification conducted in trifluoroacetic anhydride depends on the formation of cubylacyl
cation, an intermediate whose existence was demonstrated
by Xiong using NMR spectroscopy.
10.3. Higher Cubylcubanes
As the number of linked cubane units increases, that is, as
n increases, the solubility of [n]cubylcubanes plunges; at
n = 3 most are essentially insoluble in ordinary solvents.
This problem can be relieved by having moderately long
alkyl substituents on the “monomer”. Virtuani has converted 2,7-di-n-hexyl-l,4-diiodocubane
(25) into the corresponding diyl, whose polymerization, presumably initiated by
addition of 4-iodo-1-lithio- or 2,7-di-n-hexyl-l ,Cdilithiocubane, gives an oligomeric cubylcubane with a molecular
weight of approximately 10000. This corresponds to a poly-
n-G6H13
n-C6H13
Scheme 18. The key steps in the synthesis of alkylated [n]cubylcubanes. In the
first step, RLi is added slowly. R = methyl, butyl, or a long-chain substituent.
The Grignard products can be taken on in a variety of
ways. One of particular interest to us at this time is the
possibility of making liquid crystals with exceptional properties (e.g. UV transparency) from o l i g o ~ u b a n e s . [To
~ ~this
~
end Xiong has prepared materials like24 as shown in
Scheme 19.
25
n-C6H13
n-CgH 13
L
I
H
-
cubyl rod about 150 8,long containing on average 40 cubane
units !r6’1
n
flMgBr
- flcoon
con
R
R
11. Ring-Opening Reactions of Cubanes
in (CF&0)20
Cubane and many substituted cubanes are very stable
compounds, indeed remarkably so considering the extreme
strain energy of the skeleton (> 166 kcalmol- ’). There is no
symmetry-allowed, concerted pathway for ring opening.
When reagents or substituents open other paths, the molecule rearranges.
R = CH~CH~CH(CH~)CH~CH~CH~~H(CH(CH&
R
R’ = CH~CH~OCH$F&F~CF~CFZH
crystal
95%
smectic
phase B
-
145”c
.isotropic melt
Scheme 19. Synthesis of trial liquid-crystalline compounds derived from cubylcubane.
11.1. Electron-Rich Cubanes
Although cubanols can be isolated with care, they are
fragile. 4-MethylcubanolZ6, like the others, opens by way of
homoketonization to a tricyclooctenone, which in turn rearranges further to a vinylcyclobutenylketene.t661Cubyl-
It is worth noting here that esterification of the cubylcubanecarboxylic acid in Scheme 19 with the highly hindered
bicyclo[2.2.2]octan-1-01 cannot be achieved in standard
Angew. Chem.
In[. Ed. En$ 1992.31, 1421-1436
1431
amines show similar ring-opening proclivities, but their ammonium salts are quite stable. Clearly, the ring-opening process is initiated by the shift of electrons into the ring. The
tendency toward ring opening is greater when there is an
electron acceptor at C2. No cubylamine or cubanol with a
carbonyl substituent ortho to the electron donor has ever
been obtained. Similarly, 2-nitrocubylamine has never been
isolated. A simple push-pull mechanism can easily rationalize the rapid bond cleavages which account for the instability
of such compounds.
11.2. CubaneHomocubane Rearrangements
The framework of homocubane is at least 20 kcalmol-'
less strained than that of cubane. When a path is opened,
cubanes rearrange to homocubanes.
The l(9)-homocubene system contains the most twisted
C-C double bond known. If rehybridization and pyramidalization did not occur the angle between the p orbitals of the
unsaturated carbons would be 90". This is the anti-Bredt
olefin nonpareil! Its chemistry is extraordinary. 1(9)-Homocubenes rearrange to 9-homocubylidenes. These conversions are amongst the rare examples of the rearrangement of
an olefin to a ~ a r b e n e .For
~ ~ ~the
] cases at hand, using l3Clabeling, Appell et al. have shown that the rearrangement
occurs reversibly,r711and White et al. have proven that it
proceeds by the shift of a C-C bond in the homocubane
framework.1721The rearrangement can best be rationalized
by invoking a zwitterionic contributor to the olefin structure
(Scheme 20). Platz, Jones, et al. have determined the equilibrium constant for the olefin-carbene
R
11.2.1. Cubylcarbenium Ions
Great care is necessary in the isolation of cubylmethyl
alcohols such as 27 with a nucleofuge on the methyl carbon
atom. These compounds are exceedingly fragile and rearrange into the homocubyl system with exceptional ease.[67]
Scheme 20. Rearrangement of 1(9)-homocuhenes to the corresponding homocubylidenes.
The investigation of such l(9)-homocubene species is a
very active area.1731Most recently we have obtained optical
spectra of para-substituted 9-phenylhomocubenes frozen
out in 77 K glasses.[741We shall discuss these in detail elsewhere.
r27
R=H,Ph
t
11.2.3. 9-Azahomocubene
F!
The solvolysis of cubyl azide 30 may also proceed by way
of a homocubene, in this case 9-a~ahomocubene.[~~]
As yet
This probably occurs by way of a tight ion pair, as the usual
product is primarily that of internal return. The WagnerMeerwein 1,2 shift occurs even when the positively charged
carbon atom is stabilized by two phenyl groups.
I
I
I
I
11.2.2. 1(9)-Homocubenes
K.-L. Hoffmann in my group showed that cubyl phenyl
diazomethane 28 on thermolysis or photolysis is converted
there is no proof for the intermediacy of such an anti-Bredt
imine;[761the observed acetolysis product could also be accounted for by a mechanism in which solvolysis, C-C bond
migration, and nitrogen loss are linked together.
28a: R - P h
28b: ,R = H
29
to 9-phenyl-l(9)-homocubene 29a.[681Later, Jones showed
that the same reaction occurs when R = H.I6']
1432
12. Cubylmethyl Radicals
The cubylmethyl radical has been generated from cubylmethyl bromide and from the N-hydroxy-2-pyridinethione
Angew. Chem. Int. Ed. Engl. 1992, 31, 1421-1436
ester 31 of cubylacetic acid under various conditions favoring hydrogen-atom transfer to the radical. The behavior of
the radical has been studied in detail by Yip Yu Chi in my
laboratory[77a'and with N e ~ c o m b [at~ Wayne
~ ~ ' State University. Methylcubane is formed only in the presence of high
concentrations of the hydrogen donor PhSeH. Otherwise the
cubylmethyl radical rearranges. There is no evidence of a 1,2
shift to give the homocubyl system, even though this would
be extremely favorable energetically. Instead, one, two, or
three bonds of the cubane framework cleave, leading to
olefinic products. These can be accounted for by a mechanism in which ci bonds break sequentially and regioselectively; processes in which there is good overlap between the
radical orbital and that of the breaking bond are favored
(Scheme 21).[781The distribution of products depends quali-
I-\
BCH,C(O)O-N~
S
31
IhY
I
PhSeH
Treatment with Ag'
cuneane (33).[*O]
or H + induces rearrangement to
LEI
32
e
1
33
The details of such reactions were used at the time to help
unscramble the very nature of metal-induced olefin
rearrangements and olefin metathesis. Now, these early
studies are starting to pay off in a different way. The Rh'catalyzed ring-opening to syn-tricyclooctadiene is particularly useful, because this intermediate is in turn readily converted at 50-60°C into cyclooctatetraene. In my group, Chou
and Pramod have shown that this also works well with substituted cubanes, allowing much easier access to substituted
cyclooctatetraenes (e.g. 34) and oligomeric cyclooctatetraenes (e.g. 35) than previously possible (Scheme 22).[811In-
I
34
'COOCH3
Rh'
I-I
Scheme 21. Generation of the cubylmethyl radical and its reactions with various trapping reagents. The rate constants K for these reactions and the ringopening reactions are given.
0-Q
35
Scheme 22. Metal-induced ring-opening of cubane to give cyclooctatetraene.
a) Warming.
tatively on the survival time of the radical intermediates, in
other words, on the concentration and effectiveness of the
hydrogen-atom-transfer agent. From product distributions,
the rate constant for ring opening of the cubylmethyl radical
is calculated to be at least 2.7 x 101os-l; this value is substantially greater than that of any radical derived to date
from a saturated hydrocarbon system. The lifetime of cubylmethyl radical at 25 "C is about 35 p ~ . [ ~ ~ ~ ~
deed, the chemistry of cubane opens the way to new and
exciting cyclooctatetraene derivatives, for example 36.
36
13. Metal-Induced Reactions of Cubanes
More than 20 years ago Halpern, Cassar, and I showed
that various metal catalysts lead to the facile rearrangement
or ring opening of cubane at room temperature. Thus, for
example, oxidative addition to Rh' triggers opening of the
cubane frame to the syn-tricyclooctadiene system 32.[791
Angen'. Chew. Int. Ed. Engl. 1992,31,1421-1436
14. The Future
Much (but certainly nowhere near all) of the fundamental
chemistry of the cubane system is now understood. Many
new cubane derivatives and important, new high-energy sys1433
tems can be made from cubane. But what about the more
pragmatic aspects? Do cubanes have a role to play in practical chemistry? Will they prove to be important outside the
university?
r
acetylene to cubane catalytically, cubane and its derivatives
might be made cheaply. Under these circumstances cubanes
could become the fuels of choice, not only for military applications but for ordinary transportation as well. One can
1
I
b)
t
L
I think the answer is a clear “yes”. Not only are cubanes
ideally suited for drug design, high-energy cubanes are also
under consideration as explosives and propellants. Racingcar and missile manufacturers are looking to cubanes as the
fuels of the future. In other lines of applications, optically
transparent cubanes and cubylcubanes have been proposed
as building blocks for rigid liquid-crystal compounds. UVactive cubanes, for example cubyl ketones, are readily transformed photochemically into colored cyclooctatetraenes;
this conversion could be used for permanent information
storage. We are currently focusing our efforts on polymers
with cubane in the backbone or as a pendant group along a
polymer chain.[821The cubane subunits in these polymers
can be rearranged easily to cyclooctatetraenes. We expect to
convert these polycyclooctatetraenes to polyacetylenes by
way of ring-opening metathesis polymerization.1831In this
way (Scheme 23), we hope to make polyacetylenes whose
properties (e.g. stability, extrudability) are enhanced by the
chain being intrinsically part of another polymer (e.g. a
polyester or polyethylene). Of course, Scheme 23 is paper
chemistry at its most optimistic. Such propositions have innumerable problems, not the least being conformational and
configurational irregularities, cross-linking, and back-biting.
But it is time that chemists address the possibility of controlled intramolecular polymerization of “polymeric
monomers”. The early history of cubane lends credence to
the old adage that “the impossible just takes a little longer”.
Although cubane and cubane derivatives are available on
pilot-plant scale, they are still too expensive for industrial
applications. Conceptually, cubane can be regarded as a
stable oligomer of acetylene, a cheap starting material available in immense quantities. If it were possible to convert
1434
-
-
Scheme23 Hypothetical route to polymer-bound polyacetylene. a) [{ Rh(norbornadiene)Cl),]; b) metdthesis catalyst.
speculate with some pleasure on the enormous impact this
conversion would have on the way we live.
The energy balance of the conversion is good; the tetramerization is energetically favourable even when entropy
is considered. One goal we have is to discover how to achieve
this conversion and thus complete the transformation of
cubane from a laboratory curiosity to an industrial material.
I invite your participation.
The U. S. Office ofNaval Research, the U. S. Army Research,
Development and Engineering Center, the National Institutes
of Health, and the National Science Foundation have all
helped support m y work. Richard Miller at the U . S. Office of
Naval Research and Jack Alster at ARDEC have been particularly encouraging. It is a truly apleasure to acknowledge the
work of m y students and postdocs. They made our achievements possible, and they made the work itself exciting and fun.
Received: November 4, 1991 [A881 IE]
German version: Angew. Chem. 1992, 104, 1447
[l] W. Weltner, Jr., J. Am. Chem. SOC.1953, 75, 4224-4231.
(21 a ) H.-D. Martin, P. Pfohler, T. Urbanek, R. Walsh, Chem. Ber. 1983, f 16,
1415-1421; b) H:D. Martin, T. Urbanek, P. Pfohler, R. Walsh, J. Chem.
SOC.Chem. Commun. 1985,964-965; c ) W von E. Doering, W. R. Roth, R.
Breuckmann, L. Figge, H.-W. Lennartz, W.-D. Fessner, H. Prinzbach,
Chem. Ber. 1988,121, 1-9.
[3] P. E. Eaton, T. W Cole, Jr., J. Am. Chem. SOC.1964,86962-964, 3157-
3158.
(41 a) J. C. Barborak, L. Watts, R. Pettit, J. Am. Chem. SOC.1966, 88, 13281329; b) P. E. Eaton, T. W. Cole. Jr., J. Chem. SOC.Chem. Commun. 1970,
1493-1494; c) N. B. Chapman, J. M. Key, K. J Toyne, J. Org. Chem. 1970,
35. 3860-3867; d) C. G. Chin, H. W. Cuts, S. Masamune, J. Chem. SOC.
Chem. Commun. 1966, 880-881; e) G. W Griffin, A. Chaudhuri, L. W.
Reichel, R. A. Breyer, M. M. Condon, T. Elhajj, D. C. Lankin, E. D.
Anzew. Chem. Int. Ed. Engl. 1992.31, 1421 -1436
Stevens, Y. J. Li, Abstracts q/Papers, 149th National Meeting of the American Chemical Society, New Orleans, LA, Aug. 30-Sept. 4,1987;American
Chemical Society: Washington, D.C., 1987;ORGN 11. Only the synthesis
of octakis(trifluoromethy1)cubane (L. F. Pelosi and W. T. Miller, J. Am.
ChemSuc. 1976,98, 4311 -4312) and the tetrabridged cubane, propella[3,]prismane (J. Spanget-Larsen, R. Gleiter, L. A. Paquette, M. J. Carmody. C . R. Degenhardt, Theor. Chim. Acta 1978.50.145- 158)are different in concept, employing intramolecular photochemical [2 + 21cycloaddition to produce cubanes directly. Unfortunately, such cycloadditions are not applicable to the synthesis of simple cubanes; see, for example, R. Criegee, Angew. Chem. 1962,74,703-712,Angew. Chem. Int.
Ed. Engl. 1962.1, 519-527.
Cf. T.-Y. Luh, L. M. Stock, J Org. Chem. 1972.37,338-339.
P.E. Eaton. R. A. Hudson, J. Am. Chrm. SOC.1965,87,2769-2770.
D. H. R. Barton, D. Crich, W. B. Motherwell, J. Chem. Sue. Chein. Conmun. 1983,939-941.
References through late 1988 can be found in G. W. Griffin, A. P.
Marchand, Chem. Rev. 1989,89,997-1010.
L. Hedberg, K.Hedberg, P. E. Eaton, N. Nodari, A. G. Robiette, J. Am.
Chem. SUC.1991,113,1514-1517.
A. Almenningen, T.Jonvik, H.-D. Martin, T. Urbanek, L Moi. Struct.
198.5.128,239-247.
a) E. B. Fleischer. J. Am. Chem. SOC. 1964, 86, 3889-3890; b) N. L.
Allinger, P. E. Eaton, Tetrahedron Lett. 1983,24, 3697-3700.
K. D. Kybett, S. Carroll, P. Natalis, D. W. Bonnell, J. L. Margrave, J. L.
Franklin, J Anz. Chem. SUC.1966,88, 626.
D. R. Kirklin, K. L. Churney, E. S. Domalski, J. Chem. Thermodyn. 1989,
21, 3105-1113.
For comparison: the density of adamantane is approximately 1.09 gcm- ’.
J. Alster, 0 . Sandus, ARDEC, personal communication.
a) P. E. Eaton, B. K. Ravi Shankar, G. D. Price, J. J. Pluth, E. Gilbert, J.
Alster. 0. Sandus, J. Org. Chem. 1984,49, 185-186;b) P. E. Eaton, E.
Wicks, ibid. 1988,53,5353-5355.
D. S. Reddy, M. Maggini, J. Tsanaktsidis, P. E. Eaton, Tetrahedron L e l f .
1990,31,805-806.
E. W.Della, P. T. Hine, H. K. Patney, J. Org. Chem. 1977,42,2940-2941.
a)T.-Y Luh, L. M. Stock, J. Am. Chem. SOC.1974, 96, 3712-3713;
b) R. E. Dixon, A. Streitwieser, P. G. Williams, P. E. Eaton, ibid. 1991,
113. 357 -358.
For reviews see: a) H. W. Geschwend, H. R. Rodriguez, Org. Reucf. N . Y .
1979,26,1-360;b) P. Beak, V. Snieckus, Acr. Chem. Res. 1982,f5. 306312.I am particularly grateful to Prof. Snieckus for introducing me to the
power of this method.
P. E. Eaton. G. Castaldi, L Am. Chem. SUC.1985, 107,724-726.
For somewhat related examples see: G. W. Klumpp, M. Kool, A. H.
Veefkind, M. Shake], R. F. Schmitz, R e d . Trav. Chim. Pays-Bas 1983,102,
542 -543.and references therein.
a) C-H activation by an amide substituent also works for cyclopropane:
P. E. Eaton, R. G. Daniels, D. Casucci, G. T. Cunkle, P. Engel. J. Org.
Chem. 1987.52,2100-2102.b) Less strained analogues (less s character in
the C-H bonds) like adamantane or cyclobutane do not show this
phenomenon: R. Daniels, P. E. Eaton, unpublished results.
P. E Eaton. H. Higuchi, R. Millikan, Terruhedron Lett. 1987,28, 10551058.
a) P. E.Eaton, G. Cunkle, G. Marchioro, R. M. Martin, J. Am. Chem. Soc.
1987,109,948-949;
b) P.E.Eaton, R. M. Martin, J. Org. Chenz. 1988,53,
2728-2732.
A. Bashir-Hashemi, J. Am. Chem. Sor. 1988,f 10,7234-7235.
a) A. Bashir-Hashemi, P. E. Eaton, unpublished results; b) C.-X. Yang,
P. E. Eaton, unpublished results.
J. C. Bottaro, P. E. Penwell, R. J. Schmitt, J. Org. Chem. 1991,56, 13051307;see also: D.B. Reitz, S. M. Massey, ibid. 1990, 55,1375-1379.
P. E. Eaton, C.-H. Lee, Y Xiong, .
lChin. Chem. Soc. (Taipei) 1991,38,
303-306.
a) C. R. Hauser, H. G. Walker, J. Am. Chem. SUC.1947,69,295-297;
b) C. R. Hauser, F. C. Frostlick, ibid. 1949, 71, 1350-1352.
P. E. Eaton. C.-H. Lee, Y. Xiong, J. Am. Chem. Sue. 1989, 111, 80168018.
[32]Cf. R. Sanchez, W. Scott, Teirahedron Lett. 1988,29, 139-142.
[33]C.-H. Lee, Ph.D. Thesis, University of Chicago, 1990.
[34]The intermediates in many of these reactions are RMgNR, compounds.
We have dubbed them “amido Grignard reagents” and are now studying
this class of Grignard reagents and their applications in synthesis. A note
is made in passing: when these amido Grignard reagents are made from
chiral amines, chirality transfer can be (and has been) achieved in C-C
bond-forming reactions.
[35]P. E. Eaton, R. G. Daniels, D. Casucci, G. T. Cunkle, P. Engel, J. Org.
Chem. 1987,52, 2100-2102.
[36]P. E. Eaton, J. Puranik, unpublished results.
1371 For related work in other research groups see, for example: A. BashirHashemi, H. L. Ammon, C. S. Choi, J. Org. Chem. 1990, 55, 416420.
[38]P.E. Eaton. K. Pramod, R. Gilardi, J. Org. Chem. 1990, 55, 57465750.
Angew. Chem. Inr. Ed. Engl. 1992,31,1421 -1436
a) P. E. Eaton, Y. Xiong, C.-X. Yang, 1 Am. Chem. Sot.. 1990,112,32253226;b) R.M. Moriarty, S. M. Tuladhar, R. Penmasta, A. L. Awasthi.
ibid. 1990,112, 3228-3230.
D.A. Hrovat, W. T. Borden, .lAm. Chem. SOC.1990,f12,3227-3228.
a) A. J. H. Klunder, B. Zwanenburg, Tetrahedron 1972, 28, 4131-4138,
b) P. E. Eaton, S. Reddy, G. P. Sollott, J. Org. Chem. 1989.54,722-723,
P. E. Eaton, J. Tsanaktsidis, unpublished results.
R. C. Bingham, P. von R. Schleyer, J. Am. Chem. SOC.1971,93, 31893199. See also P. Muller, J. Mareda, Helv. Chim. Aria 1987, 70, 10171024.
a) E. Grunwald, S. Winstein, .
I
Am. Chem. SOC.1948, 70, 846-854;
b) T. W. Bentley, K. Roberts, J. Org. Chem. 198.5,50,4821 -4828.
There are other systems which also give cations with unexpected ease. See
for example: a) K. 8. Wiberg, V. Z. Williams, J. Am. Chem. Soc. 1967.89.
3373-3374;b) E. W. Della, P. M. W. Gill, C. H. Schiesser, J. Orf. Chrm.
1988,53,4354-4357.
For more recent treatments of cubyl cation with molecular mechanics see
a) P. Muller, D. Milin, Helv. Chim. Acfa 1991, 74, 1808-1816; b) P.
Muller, D. Milin, W. Q. Feng, R. Houriet, E. W. Della, J. Am. Chem. SUC.
1992,114,6169-6172.
The effects of electron-withdrawing groups have also been noted by
others; see [39b]and D. Kevill, M. J. D’Souza, R. M. Moriarty, S. M.
Tuladhar, R. Penmasta, A. K. Awasthi, J. Chem. SOC.Chem. Cummun.
1990,623-624.
The point has not been accepted (nor appreciated) by some authors, [39b]
and [46].
P. E. Eaton, J. Zhou, J. Am. Chem. Sue., 1992,f14,3118-3120.
For a review, see: G. Szeimies, Reorl. Infermed. (Plenum} 1983,3, 299.
For a few relevant examples, see: a) G. Szeimies, J. Harnisch, 0.
Baumgartel, 0. D. A. Hrovat, W. T. Borden, J. Am. Chem. Soc. 1977,99.
5183-5184;b) D. A. Hrovat, W. T. Borden, ibid. 1988, 110,7229-7230,
c) D. A. Hrovat, W T. Borden, ibid. 1988,f10,4710-4718;d) J. Schlfer.
G. Szeimies, Teirahedron Lert. 1988,29, 5253-5254.
P. E.Eaton, M. Maggini, J. Am. Chem. Sor. 1988,ff0,7230-7232.
1,3-Diiodocubaneis presently under investigation in my laboratory.
J. Tsanaktsidis, P. E. Eaton, Tetrahedron L e t f . 1989,30, 6967-6968,and
references therein.
P. E. Eaton, J. Tsanaktsidis, J. Am. Chem. SOC.1990. 112,876-878.
D. A. Hrovat, W. T. Borden, J. Am. Chem. Sue. 1990,112, 875-876.
K. Hassenruck, J. G. Radziszewski, V. Balaji, G. S. Murthy, A. J. McKinley, D. E. David, V. M. Lynch, H.-D. Martin, J. Michl, J. Am. Chem. Suc.
1990,112, 873-874.
Nevertheless, it is amusing to point out that 20 (= 2 0 ) is related to
[2.2.2]propellaneA, a compound synthesized many years ago in my laboratory The formation of single bonds in highly strained molecules by
through-space dehalogenation has a checkered history; see footnote 2 in
ref. [54]for a brief discussion. P. E. Eaton, G. H. Temme, J. Am. Chem.
SOC.1973,95, 7508-7510.
A
20
20’
R. Gilardi, M. Maggini, P. E. Eaton, J. Am. Chem. SUC.1988,110,72327234;see also [56].
0. Ermer, J. Lex, Angew. Chem. 1987,99,455-456;
Angew. Chem. Int. Ed.
Engl. 1987,25,447-449.
R. A. Alden, J. Kraut. T. G. H. Taylor, J. Am. Chem. SOC.1968,90,74-82.
I am grateful to D. S. Reddy and A. Basir-Hashemi for informing me of
the results of their initial exploration of this reaction.
P. E. Eaton, K. Pramod, unpublished results.
P. E. Eaton, Y. Xiong, unpublished results.
This work is being done in collaboration with D. Walba, University of
Colorado, and M. Wand, Displaytech, Inc.
P. E. Eaton, M. Virtuani, K. Pramod, unpublished results. The molecular
weight assigned assumes the applicability of the universal calibration
parameter, see 2. Grubisic, P. Rempp, H. Benoit, Polymer Lett. 1967,5,
753.
R. E. Hormann, Ph.D. Dissertation, University of Chicago, 1987.
a) T. W Cole, Jr., Ph.D. Dissertation, University of Chicago, 1966;b) T.-Y
Luh, Ph.D. Dissertation, University of Chicago, 1974.
P. E. Eaton, K:L. Hoffmann, J. Am. Chem. SOC. 1987, 109, 52855286.
M. Jones, Jr., N. Chen, J. Phys. Org. Chem. 1988,1,305; Tetrahedron Lett.
1989,30,6969-6972.
See footnotes 13, 14, 15 in ref. [68]and also a) T. J. Barton, M.-H. Yeh.
Tetrahedron Left. 1987,28, 6421-6424;b) B. Halton, J. H. Bridle, E. G.
Lovett, ibid. 1990,31, 1313-1314.
P.E. Eaton, R. B. Appell, J. Am. Chem. Sor. 1990,112, 4055-4057.
1435
1721 P. E. Eaton, A. J. White, J. Org. Chem. 1990, 55, 1321-1323.
1731 See, for example: N. Chen. M. Jones, Jr., W. W. White, M. S. Platz, J. Am.
Ckem. Sac. 1991, 113, 4981-4992; P. E. Eaton, R. Appell, W. R. White,
M. S . Platz, unpublished results.
[74] R. B. Appell, Ph.D. Dissertation, University of Chicago, 1990.
[75] P. E. Eaton, A . M . Fisher, R. E. Hormann, Synkrr 1990, 737-738.
[76] Photolysis of cubyl azide leads to the formation of homopentaprismanecarbonitrile. Matrix isolation studies proved insufficient to demonstrate
the intermediacy of azahomocubene: P. E. Eaton, R. E. Hormann, J. A m .
Chem. SOC.1987, 109, 1268-1269.
1771 a) P. E. Eaton, Y.C. Yip, J. Am. Chem. SOC.1991.113, 7692-7697; b) S.Y Choi, P. E. Eaton. M. Newcomb, Y.C. Yip, &id. 1992, 114, 63266329.
1781 For earlier examples, see: a) M. Castaing, M. Pereyre, M. Ratier, P. M.
1436
1791
[XO]
[Xl]
[82]
1831
Blum, A. G. Davies, J. Ckem. SOC.1979, 287; b) A. L. J. Beckwith, G.
Moad, J. Am. Chem. SOC.1980, 92, 1083, and references therein; c) K. U.
Ingold. B. Maillard, J. C. Walton, ibid. 1981, 93, 970, and references
therein.
L. Cassar, P. E. Eaton, J. Halpern, J. Am. Chem. Sor. 1970,92,3515-3518.
a) L. Cassar, P. E. Eaton, J. Halpern, J. Am. Chem. Sac. 1970, 92, 63666368; b) J. E. Byrd, L. Cassar, P. E. Eaton, J. Halpern, L Chem. Sor.
Chem. Commun. 1971, 40-41.
P. E. Eaton, C:T. Chou, K. Pramod, unpublished results.
a) P. E. Eaton, S. Giacobbe, S. Cohen, Poster Presentation, Arnoco Oil
Company, October, 1991; b) P. E. Eaton, C. T. Chou, S . Giacobbe, L.
Zhu, unpublished results.
See, for example: F. L.Klavetter, R. H. Grubbs, J. Am. Ckem. Sac. 1988,
110, 7807-7813.
Angew. Chem. Inc. Ed. Engl. 1992,Si. 1421-1436
Документ
Категория
Без категории
Просмотров
5
Размер файла
1 313 Кб
Теги
chemistry, 1990s, starting, cubanes, material, century, new
1/--страниц
Пожаловаться на содержимое документа