close

Вход

Забыли?

вход по аккаунту

?

Bredt Compounds and the Bredt Rule.

код для вставкиСкачать
[70] F . A. Neugebauer, Tetrahedron 26,4843 (1970).
[71] F. A . Neugebauer, Angew. Chem. 81, 496 (1969); Angew. Chem.
internat. Edit. 8, 520 (1969).
[72] F. A. Neugebauer and H . Fischer, Chem. Ber. 106, 1589 (1973);
cf. D.Jerchel and U! Edlrr, Chem. Ber. 88, 1284 (1955).
[73] N . Azuma, K . Mukai, and K . Jskizu, Bull. Chem. SOC. lap. 43,
3960(1970).
[74] F. A . Neugebauer and G . A . Russrtl, J. Org. Chem. 33, 2744 (1968).
[75] F. A . Neugebauer, Chem. Ber. 102, 1339 (1969).
1761 K . H . Hausser, Z. Naturforsch. 1I a, 20 (1956).
Bredt Compounds and the Bredt Rule
By Gert Kobrichl*]
Dedicated to Professor Theodoi- Wieland on the occasion of his 60th birthday
The Bredt rule, which proved useful in earlier decades mainly as an exclusion criterion
in the assignment of structures, has in recent years become the stimulus for efforts to
synthesize strained bicyclic systems with a bridgehead double bond. On the basis of the
ring-strain forces operating in such systems, an extended Bredt rule is discussed. The “Bredt
compounds” to which this rule applies, which include bicyclic systems not previously considered from this point of view (e.g. bridged methylenecyclopropanes), form a clearly circumscribed, interesting field within the chemistry of highly-strained hydrocarbons with ring
unsaturation.
1. Introduction
According to the rule formulated by Bredt and called
after him, “in systems of the camphane and pinane series
and related compounds ( I ) and (Z), the branching points
A and B of the carbon bridges (the bridgeheads) cannot
be involved in a carbon double bond‘“‘’.
B
7,
1
An experimental observation that led Bredt to this realization shortly after the turn of the century will be outlined
here[’]. Firstly, the attempted dehydrobromination of the
anhydride ( 3 ) was unsuccessful,whereas the corresponding
dicarboxylic ester could be smoothly dehydrobrominated;
secondly,the unsaturated dicarboxylic acid ( 5 ) was found
to be incapable of forming its internal anhydride ( 4 ) ,
whereas saturated analogs underwent this cyclization very
readily. The resistance to the formation of ( 4 j and related
bicyclic structures was correctly attributed by Bredt to
the ring strain that would necessarily result from the distorted geometry of the carbon double bond at the bridgehead in comparison with normal olefins.
[*] Prof. Dr. G. Kobrich
Institut fur Organische Chemie der Technischen Universitat
3 Hannover, Schneiderberg 1 B (Germany)
Based on a lecture given in Regensburg on July 3, 1972
464
The correctness of Bredt’s hypothesis was soon shown
very impressively by the fact that it was found necessary
to revise all earlier structural assignments contrary to his
p~stulate”~.
The Bredt rule therefore initially served, and
still serves, as an exclusion criterion in the determination
of constitutions of unsaturated bicyclic compounds. This
aspect, which is now mainly of historical interest, will
not be considered further here, particularly since it has
been presented in great detail by F a w ~ e t t [in
~ ]a survey
of the literature published up to 1950.
Another aspect has proved to be interesting and important
for the f u t ~ r e [ ~If, ~the
l . ring strain prevents the formation
of a double bond at the bridgehead in the bicyclic structures
( I ) and (Zj, it will certainly also do so in compounds
with smaller rings, e.9. in the bicyclo[2.l.l]hexane system;
however, this rule will not be absolutely valid for the
higher homologs of ( 1 ) and (Z), since with a sufficiently
large number of atoms in the ring it is possible to construct
completely strain-free molecular models of bridgehead olefins. This raises the question how far the Bredt rule is
valid for the higher homologs of ( I ) and (Z), or in other
Angew. Chem. internat. Edit.
Vol. 12 (1973) 1 No. 6
words, the ring size beyond which it is no longer valid.
When considered in this way, the rule changes from a
fact to a problem, and from an aid in the investigation
of structures to a subject for investigation, whose current
interest is shown by a number of very recent studies.
It would admittedly be inappropriate to consider this problem in isolation. It forms part of the broader chemistry
of strained ring systems, with which it is closely interwoven.
The aim of the following discussion is to fit the Bredt
rule into its place in this wider context.
2. Isolable Bridgehead Olefins
The range of validity of the Bredt rule, in the sense of
our problem, can be defined only by negative evidence.
The successful preparation of a bicyclo[x.y.z]alk-I-ene
shows that the rule is not valid for this compound, though
failure to prepare this compound does not necessarily
indicate the validity of the rule. The information derivable
from an experiment is thus crucially dependent on the
appropriate choice of experimental conditions, particularly
the "correct" synthetic pathway. A highly strained compound will be formed only if a sufficiently energy-rich
precursor that reacts irreversibly and in only one direction
is available, and its isolation will be successful only if
no subsequent reactions occur.
The first attempts to establish a boundary were made
in 1948/49 by Prelog et al!'] in the intramolecular aldol
condensation of the compounds (6). The compounds (7)
with a bridgehead double bond were formed smoothly
from the homologs with n>5; with n=5, (8) was formed
in addition to (7) in an alternative reaction, and with
n < 5 (7) was no longer obtained.
bridge containing the bridgehead double bond will be
indicated by a bar. The compound ( 7 ) , n=5, for which
the Bredt rule is no longer valid according to Prelog et
al., is thus a [5.3.I] system with S=9. The ability of even
smaller bicyclic structures with a bridgehead double bond,
such as the C5.3.11 compound (13), with S=7, to exist
at least for a short time was suggested by the smooth
decarboxylation of bicyclic P-keto acids such as (9)[81
and
According to the current view on the decarboxylation
mechanism of P-keto acids["', this reaction should proceed
via enols having the ring skeleton ( I I). A similar conclusion can be drawn from the triple H/D exchange in the
ketone (12) in basic media[111.In 1967, the research groups
of Marshall[' and Wisernan['3a1succeeded in preparing
the stable bicyclo[3.3.l]non-l-ene ( 1 3 ) by the three paths
outlined below. The aza analog ( 1 3 a ) is now also
known[l41.
'OH
The isomeric [4.2.1] and La.2.11 systems ( 1 4 ) and (16)
were also synthesized shortly afterwards. Hofmann elimination in ( I S ) , with kinetic product control, gave the
(presumably) thermodynamically less stable ( I 4 ) in higher
yield than (16)l' b].
'8,
For purposes of comparison, it is useful to denote the
sum x + y + z of the numbers of bridge atoms in a bicyclo[x.y.z]alk-1-ene by S, as was done by F a w ~ e t t [ ~In
I.
the following discussion, moreover, the number for the
Angew. Chem. internat. Edit.
1 Vol. 12 (1973) N o . 6
3. Relations between Structure and Ring Strain
The hydrocarbons ( 1 3), ( 1 4 ) , and ( 1 6) at present mark
the lower limit for stable bridgehead olefins. This admittedly does not mean that compounds having S = 7 in
general lie outside the range of validity of the Bredt rule,
since the number of bridge atoms S gives only an incomplete picture of the ring strain. It can be used for predictions
regarding strain and stability only when one is considering
similar (homologous) bicyclic structures. Bridgehead olefins might even possibly be more stable than bicyclic systems with different structures and with a higher S. Another
determining factor for a given S is the size of the individual
bridges. The stability also depends on whether, for a fixed
465
ring skeleton, the bridgehead double bond is situated in
a larger or in a smaller bridge. The following rules, which
admittedly d o not yet allow comparison in every case,
can be deduced for bicyclic structures with the aid of
models:
Rule A. For homologs with different S values, the ring
strain varies inversely with S41.
Rule B. For a given S , the ring strain varies inversely
with the size of the larger of the two rings with respect
to which the bridgehead double bond is endocycli~['~]
(for explanations and reasons see Section 7).
Rule C . For a given bicyclic ring skeleton, the ring strain
varies inversely with the size of the bridge containing
the bridgehead double bond.
undergoes silver-salt catalyzed skeletal rearrangement to
the cyclononene compound ( 2 7 ) , the [4.3.T] system ( 2 6 )
being a suspected
We encounter a particularly interesting bridgehead olefin
in adamantene ( 2 9 ) which turns out to be a derivative
of the [3.3.i] system (19) when drawn in the distorted
fashion chosen for formula (30), but with an even higher
ring strain owing to the additional methylene group. Its
intermediacy in the dehalogenation of 1,Zdiiodo (and 1bromo-2-icdo-)adamantane ( 2 8 ) by organolithium compounds is apparent from the almost quantitative formation
of a hydrocarbon C20H28[471r
which is presumably a headto-tail dimer0f(29)[~~"].
( 2 9 ) is assumed to have a "double
bond" with completely decoupled x
b!
Rule C explains, for example, the greater stability of (16)
in comparison with ( 1 4 ) . It follows from rules B and
C, for example, that ( 1 5 ) does not give the least stable
C4.2.71 system (I 7 ) , whose lower formation tendency
to decarboxylate the
is also reflected in the
P-keto acid (18).
Thus no isolable [ x . y . i ] compounds have so far been
obtained with S=7; the bridgehead olefins of this structure
under current investigation belong to the unstable intermediates to be discussed below.
According to rule C, among the as yet unknown bicyclic
structures (19) to (21), (19) should be less stable than
( 1 3 ) , and ( 2 1 ) should be less stable than (20). The investigation of these systems with S = 7 should be interesting
and important, since they presumably lie on the borderline
between stable and unstable compounds as defined in
Section 6.
Initial reports of such investigations have appeared in
very recent publications. Thus the [4.2.i] compound ( 2 3 )
is formed as a short-lived intermediate by cyclopropyl-ally1
rearrangement of dichlorocyclopropane ( 2 2 ) at about
25 "C, as shown by the isolation of the head-to-head dimer
( 2 4 ) whose structure has been elucidated by X-ray methods
(mirror symmetry)'46a].The ring strain in (23) is estimated
as 2&22 kcal/mol. The dichloride (25a), a higher homolog of ( 2 2 ) ,is thermally stable. However, the corresponding
dibromide (25 b ) (and likewise also the dihydro derivative)
4. Unstable Compounds with S < 7
According to the current view, bridgehead olefins with
S<7 are capable of existing, but are unstable. Our as
yet very incomplete knowledge is based on the circumstantial evidence that is usual for short-lived intermediates,
and is subject to the uncertainties associated with this
method.
This is true e.g. for the above-mentioned decarboxylation
of P-keto acids with carboxyl groups attached to the bridgehead carbon atom. The fact that decarboxylations occur
with difficulty if at all in systems with S<6'16-191can
be safely interpreted as resistance to the formation of
the corresponding ring-strained enols. On the other hand,
it is very doubtful whether the normal decarboxylation
mechanism via enols [see formulas (10) and ( I J ) ] is
followed under forcing conditions, as in the thermolyses
of the P-keto acids (31), (32)[18J and (33)[191 between
260 and 320°C.
A variant has been proposedr'81 (and criticized[13".201 1,
in which the bridgehead atom retains its sp3 hybridization
in the course of the elimination of COz. According to
the decarboxylation of optically active
Buchanan et
(33), contrary to expectation, proceeds with complete race466
Angew. Chem. internat. Edit. / Vol. 12 (1973) /
No. 6
mization. It could therefore involve an initial ring cleavage,
followed by the normal decarboxylation and a recyclization. It is significant that the rnonocarboxylic acid formed
from (33) is not decarboxylated furtherr2’!
v
S
The decarboxylation of small bicyclic P-keto acids
obviously does not allow reliable deductions concerning
the existence or non-existence of ring-strained bridgehead
olefins, owing to its unconfirmed mechanistic premises.
Nevertheless, the recently accomplished base-catalyzed
H/D exchange at the bridgeheads of the ketones (34)[48a1
and (35)[48b] that are adjacent to the carbonyl groups
provides some indication of the existence of enolates of
type (36), however short-lived they may be.
Azabicycloolefins with S = 6 have been variously postulated
as intermediates. Thus the thermolysis of 0-tosyl-2-quinuclidinyl phenyl ketoxime proceeds partly as a fragmentation viu the carbenium ion (38a) and is faster than that
of the homomorphous alicyclic compound, which decomposes exclusively by Beckmann rearrangement. The greater
formation tendency of ( 3 8 a ) can be explained by resonance
stabilization involving the bridgehead-irnmoniurn mesomeric structure (38b)[”].
The azabicyclic compounds (39) and (40) formed in
roughly equal amounts on photolysis of l-azidonorbornane in methanol could be derived from the primary
ring-expansion products ( 4 1 ) and ( 4 2 ) by addition of
methanol. However, the authors[””] favor a synchronous
process that avoids these energy-rich intermediates. The
N-protonated ( 4 2 ) is nevertheless postulated as an intermediate in the reduction of (40) with LiAlH, [to (40)
with H instead of OCH3][22a1.The unsaturated lactam
(43) is a plausible intermediate in the oxidation of the
corresponding saturated lactarn with lead tetraacetate[22b1.
Evidence has recently also been obtained for the intermediacy of bicyclo[3.2.1]oct-l-ene~13c~:
thermolysis of the
thiocarbonate
2,4-dioxatricyclo[6.2.1 .O ’. S]undecaneAnyew. Chem. inrernat. Edit. 1 Vol. 12 (1973) 1 N o . 6
3-thione with trialkyl phosphite in the presence of 1,3diphenylisobenzofuran yields 62% of a mixture whose spectral data are compatible with the structure of the expected
Diels-Alder adduct (mixture of stereoisomers). Hofmann
elimination with the structurally analogous quaternary
ammonium hydroxidegives 0.3% of a product which could
also contain the adduct of the [3.2.1]bicycloolefin.
For these reasons and on the basis of the following observations, therefore, it may be assumed that bicyclic systems
with S=6 and a double bond at the bridgehead can exist,
though there is as yet no satisfactory experimental proof.
It should be stressed that conclusive proof of this would
not be contrary to the Bredt rule in its original version,
which was formulated for systems with S = 5. It is interesting t o note, however, that two “forbidden” 1-norbornenes
(S = 5 ) have recently been detected as short-lived interrnediates. In 1965, Tatlow et al. observed the elimination
of LiF from 1-lithioperfluoronorbornane ( 4 4 ) to give perfluorinated 1-norbornene ( 4 5 ) , which was characterized
mainly by its diene reaction with furan to form the two
possible stereoisomers of the cycloadduct (46)[23a1.If the
halogen at the bridgehead of ( 4 4 ) is replaced by H, how-
1441
(4.5)
1461
ever, the p elimination does not take p l a ~ e f ~Because
~~l.
of this surprising substituent effect, one cannot deduce
the existence of the unsubstituted compound (48) from
that of ( 4 5 ) . The existence of ( 4 8 ) was demonstrated
in 1971 by Keese and E.-P. K r e b ~ [ ’ ~Dehalogenation
]:
of the vicinal dihalides (47) with butyllithium in the presence of furan results in the formation of the cycloadduct
( 4 9 ) . The I-norbornene intermediate (48) is confirmed
by the discovery that the two stereoisomers of ( 4 9 ) are
always formed in a ratio of 1 :4, regardless of the nature
of the halogen and of the organometallic agent.
There is so far no evidence of the existence of the isomeric
norbornene (50). Its relatively (according to rules B and
C) and absolutely low formation tendency is indicated
on the one hand by the apparently exclusive formation
467
X
centers. These deformations combine to give a resultant
that also influences the degree of overlap of the two p
orbitals.
of ( 4 5 ) from ( 4 4 ) and on the other by the unusual thermal
stability of the Grignard compound ( 5 1 ) (no elimination
at 80"C)r251.
Little is known at present about the effects of these deformation forces on the nature of the "double bond". The distortion of the 7~ bond is presumably accompanied by a change
in hybridization, with admixture of s components into
the p orbitals and an increase in the p character of the
(3 bonds. For trans-cyclooctene, whose relation to bridgehead alkenes will be discussed in Section 7, the configuration shown in formula ( 5 6 ) , with the bonds bent downwards and the orbital lobes on both carbon atoms extended
in the opposite direction, has been proposed[26! In bridgehead alkenes with non-equivalent olefinic carbon atoms,
this change of hybridization should be greater for the
bridgehead atom, which is more exposed to the shearing
forces.
5. Causes of the Ring Strain
The instability of the bicyclic systems to which the Bredt
rule is applicable is a result of the increased ring strain
in relation to saturated compounds having the same structure, as mentioned earlier. The additional strain is due
partly to the widening of tetrahedral angles to 120°, and
mainly to the fact that a double bond does not exhibit
the free rotation of 0 bonds. The dihedral angles formed
by the single bonds from the bridgehead atom with the
valences from the other atom of the double bond are
therefore no longer unimportant. The dihedral angles of
0" and 180" in the planar olefins cannot be achieved at
bridgeheads of small bicyclic systems, since three of the
four bonds from the ends of the double bond, as parts
of the bicyclic system, curve away in the same direction,
i. e. toward the other bridgehead [see formula (5211. The
two olefinic carbon atoms cannot therefore achieve the
usual coplanar arrangement with the four atoms bonded
to them, nor can they form the normal bond angles of
120".The reasons for the ring strain are therefore as follows:
1. The most important and most characteristic reason
is the twisting of the n bond about the axis joining the
two olefinic carbon atoms['].
This is a consequence of the fact that the four bonding
partners of the olefinic carbon atoms cannot lie in the
same plane, and optimum overlap of the two p orbitals
is therefore prevented. The extreme case, with a twisting
angle of 90', would be a biradical or C-ylide [formula
(53)l.
2. Planar [formula ( 5 4 ) ] and nonplanar {formuIa ( S S ) ]
deformations of the (3 bonds emerging from the unsaturated
468
According to preliminary model calculations by Keese
and Krebs, the configuration shown in (57), with a roughly
sp3-hybridized bridgehead atom and a p orbital on the
neighboring atom, is in fact energetically more favorable
than ( 5 6 ) for 1-norbornene (48). Its n-bond order is only
about half of that of ethylene, and it carries an excess
charge on the bridgehead atom[24b].
The question arises whether the extreme case is not more
adequately represented by a dipolar (C-ylide) structure
e,
,c--c,
e,
than by a diradical. No confirmatory experimental evidence
has yet been obtained for this assumption. In polar additions to bicycIo[3.3. llnon- I-ene (131, the nucleophile,
contrary to the above model and in agreement with the
Markownikoff rule, always adds to the bridgehead
atom[". 13a1;however, this compound ismuch less strained,
and is therefore possibly atypical. The spontaneous cyclodimerizations of unstable bridgehead olefins could provide
more reliable information: assuming comparable steric
conditions, head-to-tail cycloaddition should take precedence over the head-to-head process. However, the few
results obtained so far indicate occurrence of both types.
6. Bredt Compounds
There are no empirical parameters available at present
for the quantitative correlation of structural effects and
ring strain. The Bredt rule is thus a qualitative statement
with fluid boundaries. It would be unreasonable (and it
Angew. Chem. internat. Edit. 1 Vol. 12 ( 1 9 7 3 ) 1 N o . 6
is not usual in the literature) to regard it in its original
version (Section 1 ) as a strict prohibition of existence for
certain compounds. Exceptions were mentioned in Section
5, and future investigations will undoubtedly show that
other “forbidden” bicyclic systems can exist, so that the
Bredt rule would ultimately become meaningless, despite
its inherent value. The objection that the “forbidden” compounds d o not possess a “true” double bond, but are
biradicals or betaines, would be too sophistic an apology.
It seems much better to interpret the Bredt rule as characterizing an unstable rather than a non-existent class of
bicyclic compounds. It is then possible and meaningful
to refer to the bicyclic systems to which it is applicable
as “Bredt compounds” (see also Scheme 1 in Section 10)
and to define them as follows: “Bredt compounds are
bicyclic (and polycyclic) systems (alicycles and heterocycles)
that, in addition to a strained o-bond skeleton, have a
twisted R bond at a bridgehead, and purely because of
this ring strain, in contrast with compounds having the
same structure but without this R bond, are unstable at
room temperature.”
7. Comparison with Strained trwns-Cycloalkenes
The above definition of Bredt compounds is based primarily on ring-strain rather than structural criteria. It is thus
more comprehensive than the original Bredt rule, even
in its extended form given in the introduction. T o illustrate
the consequences, let us consider some classes of ring-unsaturated compounds whose unstable members, in part, are
not Bredt compounds by definition, and in part have until
now been traditionally regarded as not being subject to
the Bredt rule.
The relationship of the Bredt bicyclic systems with the
rrarzs-cycloalkenes, which was recognized as early as 1933
by Ehc>l[“1, is particularly striking. According to A!the ring strain in trans-cyclooctene ( 5 8 ) . the
smallest stable memberf2’1, is due to the same deformation
forces that occur in Bredt compounds according to Section
5 , as can be seen from molecular models. In fact, bicyclo[3.3.l]non-l-ene ( 1 3 ) (Section 2), the smallest stable
bridgehead olefin, may be regarded as a bridged frans-cyclooctene, and can be constructed from the latter in models
without much additional strain [formulas ( 5 8 ) and
(59)]“3”1. According to thermochemical data, ( 1 3 ) has
a ring strain of 12 kcal/mol, as compared with 9.2 kcal/mol
in tran.s-cycIooctene[2yl.The calculated heat of hydrogenation of ( 1 3 ) is 12.85 kcal/mol less, i.e. its ring strain is
greater by this amount, than that of the A 2
The instability of bicycloolefins smaller than ( 2 3 ) is paralleied by the instability of fruns-cycI~heptene[~~~,
whose
calculated ring strain is about 10 kcal/mol higher than
hl.
that of trui~s-cyclooctener~~
A double bond at the bridgehead of a bicyclic system
is always exocyclic with respect to one of the three rings
Anyew. Chem. inremat. Edit.
Vol. 12 (1973) f N o . 6
and endocyclic with respect to the other two. According
to Wisrman[131,
the number of atoms in the larger of the
two rings with the endocyclic double bond provides information about the relative stability of the bicycloalkene
(rule B, Section 3). This explains e. y. the greater stability
of the [?.3.1] and C4.2.11 systems ( 1 3 ) and ( 1 6 ) in comparison with the 13.3.11 and [4.2.i] isomers ( 1 9 ) and ( 1 7 ) .
since the largest ring with an endocyclic double bond
is cyclooctene in the first two compounds as opposed
to cyclohexene and cycloheptene, respectively, in the other
two. The applicability of this rule is admittedly limited,
since it does not allow comparison of isomers such as
(14)and(16)or(20)and ( 2 1 ) ([Y.y.z] and [x.r.z) systems
in general) that differ only in the position of the double
bond in the largest unsaturated ring. In this case rule
C provides supplementary information.
8. Systems Containing a Zero Bridge
[X.y.O] systems, i. e. bicyclic systems with a zero bridge,
were always tacitly or expressly[41excluded from the problem of the Bredt rule in the past. This restriction does
not exist in our definition of Bredt compounds. If an
unstable [T.y.O] system exhibits the strain characteristics
mentioned, it can be classified as a Bredt compound; otherwise it cannot. To establish a limit, let us consider some
strained bicyclic systems: in order to determine the causes
of their ring strain, we shall again use their imaginary
derivation from unsaturated monocycles.
A revealing limiting case from the standpoint of the last
section is the strained hydrocarbon ( 6 0 ) , which was prepared by Wrinshenker and Grwne in 1968f3’1(its 5-chloro
derivative was known even earlier as an
It may be regarded as a doubly bridged trans-cyclooctene,
but on symmetry grounds, only non-planar deformation
forces act on its double bond (see Section 5 ) ; it thus lacks
the twisting of the double bond that is typical of Bredt
compounds. A significant feature is the direct linkage of
two bridgeheads by a double bond.
The same is true of IJ-bridged cyclopropenes (62) and
cyclobutenes ( 6 3 ) , which can be derived from the corresponding cycloalkynes (61 ) by expansion of the R bond,
which lies in the plane of the ring, to a three-membered
or four-membered ring. The smallest stable cycloalkyne
469
is the sterically shielded 3,3,7,7-tetramethylcycloheptyne
[skeleton: ( 6 2 ) , n = 5]'331;the cyclopropenes ( 6 2 ) become
unstable on transition from n = 5 to n =4[341,and the cyclobutenes ( 6 3 ) on transition from n=3[351 to n=2[36'.
thecompound ( 6 9 ) , which is identical with ( 6 7 ) , for n=2.
This should have a ring strain roughly comparable with
that of ( 6 8 ) , but the twisting strain is slightly reduced
in favor of the deformation forces (see Section 5). Thus
In the cyclobutatrienes (64), with 1.2,3-cyclodecatriene
[ ( 6 4 ) . n=6] as the smallest stable
expansion of the x bond, which is situated in the plane of
the ring, leads to the bridged bismethylenecyclopropanes
( 6 5 ) , which should be isolable with n = 5 or even 4 because
of the associated bending of the central CF bond.
IhS,
ffi9,
if an [X.l.l] or [ x . i . l ] system is a Bredt compound, the
[?.1.0] or [x.i.O] system with one carbon atom less may
be assumed to be a Bredt compound too. We note as
a selection rule, however, that a one-carbon bridge containing a double bond cannot be removed, since this would
give a bicyclic system with a double bond between the
two bridgeheads.
In small-ring cycloalkynes and cyclobutatrienes, the x
bonds, whose orbitals lie in the plane of the ring, are
subject to planar deformation forces, but not to twisting
strains. The same is true of the bicyclic systems (62),
( 6 3 ) , and ( 6 5 ) , as is confirmed by molecular models.
Unstable bicyclic systems with a zero bridge are thus
not Bredt compounds if the double bond is also the zero
bridge or if the same bridge contains another double bond
at the other bridgehead.
The situation is different in bicyclic systems that can be
thought of as derived from cycloallenes ( 6 6 ) . The smallest
compound of this type that is stable at room temperature
is I ,'-cyclononadiene [ ( 6 6 ) , n =6y3*]; lower homologs
with n=3 to 5 are
but the homolog with
n = 5 has been detected directly by spectroscopy at low
Cycloallenes present another, formerly overlooked analogy
between the ring strain of unstable monocyclic systems
and that of Bredt compounds. Short methylene bridges
cause not only planar and non-planar deformations of
the two double bonds, but also twisting of the x bonds,
since they distort the end valences of the allene from
the natural dihedral angle of 90- toward a coplanar arrangement. If one x bond of the allene is expanded to
a three-membered ring, we obtain a bridged methylenecyclopropane (67). The same forces typical of Bredt compounds that act in the cycloallene ( 6 6 ) also act in the
methylenecyclopropane, but the ring strain for a given
ring size is less, since the three linearly arranged carbon
atoms in ( 6 6 ) now form an angle of about 150, and
their end valences form a dihedral angle of less than 90".
This relationship with "classical" Bredt compounds can
also be demonstrated as follows. If a carbon atom bonded
to two bridgeheads is cut out of bicyclo[3.l.l]hept-I-ene
( 6 8 ) , which certainly is a Bredt compound, we obtain
470
9. Bridged Methylenecyclopropanes as a
Test Case for the Bredt Rule
According to the above argument, bridged methylenecyclopropanes ( 6 7 ) can be used as a basis for statements regarding the range of validity of the Bredt rule. Compounds
( 6 7 ) with n > 6 are obtainable by carbene additions to
cycloallenes ( 6 6 ) , and are stable[371.According to Gardnrr
et
lower homologs are formed on dehydrobromination of bridged bromocyclopropanes (70) with strong
bases.
c
However, only bicyclo[6.1.0]non-l-ene [ / 6 7 ) , n= 51 can
actually be isolated; the didehydro derivative ( 7 2 ) can
be prepared analogously by double dehydrochlorination
of 9,9-dichlorobicyclo[6.1 .0]nonane14 'I. As was mentioned earlier, these two compounds may be compared
with bicyclo[6.1. lldec- 1-ene, S = 8, and are therefore stable
asexpected. The bicycles ( 6 7 ) with n 45 that are of interest
to our discussion, on the other hand, cannot be isolated
in basic media, because of their rearrangement to the
less strained A2 isomers f 73)r4'"1; more highly unsaturated
derivatives undergo further additions and skeletal rearb.421. Since the formation conditions favor
rangement~'~'
isomerizations, no conclusions can be drawn from these
results concerning the thermal stability of the bridged
methylenecyclopropanes ( 6 7 ) with n< 5. The same is true
for the bridged cyclopropenes ( 71), whose unsaturated
member (74) was postulated as an intermediate in the
phenylcarbene rearrangement[431.
Literaturedata allow the following assessment. The bridged
methylenecyclopropanes ( 6 7 ) are formally links between
Angew. Chem. internat. Edit. J Vol. 12 (1973) J N o . 6
the cycloallenes ( 6 6 ) and the tricyclo[x. 1.0.0’~3]alkenes
( 7 5 ) . The former are stable from the nine-membered ring
on (see above), and the latter from the five-membered
ring on [( 7 5 ) ,n = 2][43].It should therefore also be possible
product at -40 C is the tricyclic compound ( 8 5 ) (the
spectroscopic data are also in accord with a structure
having two methyl groups at neighboring bridgehead
atoms), which is probably formed via the trimethylenemethane intermediate (84)[44c1.
to isolate the compound ( 6 7 ) with a seven-membered
ring (n=4). ( 6 9 ) , which contains two bridge atoms less
than ( 6 7 ) , n = 4 , is clearly unstable according to the above
arguments, and the link ( 7 6 ) is thus an interesting test
case.
Starting with suitably substituted chloroolefins of the type
( 7 7 ) , we checked these predictions for the intramolecular
cyclopropanation of unsaturated carbenes (78) as represented by the following scheme, which, having a high
driving force, proceeds under mild conditions and without
subsequent i s ~ m e r i z a t i o n f ~ ~ ~ .
I!i
i S i;
is41
The limits of the cyclization 78)-(67) are reached with
the trimethyl derivative of (78) with n = l , which reacts
intermolecularly instead of undergoing the intramolecular
reaction to form the enormously strained bicyclo[2.1 .O]pent- I-ene ( 8 1 ) ( “ h ~ u s e n e ” ) [ ~ ~ ] .
(77,
7s,
(67,
In this way a (substituted[*])chloroolefin ( 7 7 ) with n = 4
gave a stable hydrocarbon, to which the constitution ( 7 9 )
was assigned[44a1. From a lower homolog with n=3, it
was possible to obtain the compound ( 8 0 ) , whose double
bond at the bridgehead has been ~ o n f i r r n e d [ ~
The
~ ~comJ.
pound (80) decomposes at room temperature with a halflife of about 70 hours, and can thus be classified as a
Bredt compound. The next lower homolog (821, as
expected, could not be isolated. Its transient existence
has been deduced from the formation of dimeric hydrocarbons C I 8 H Z 8The
: pentacyclic system ( 8 3 ) , whose structure has been established by X-ray a n a l y ~ i s [ ~ ~is‘ I ,the
main product 6 - 50°C and obviously results from a headto-head t’xo addition of two molecules of (82). The major
p] The methyl groups prevent elimination to form an acetylene and
facilitate the cyclopropanation. The additional double bond in ( 6 7 )
prevents intramolecular CH insertion of the intermediate carbenoid,
which always occurs readily if the formation of a cyclopentene is posslble
[Ma, 44e]:
On the basis of the above results and arguments, it can
be predicted for the homologous series of bicycIo[x. 1 .I]alkI-enes that (86) (S=6) and (87) ( S = 5 ) are compounds
that can exist with moderate and short lifetimes respectively; the higher homolog (20) (S=7) might be stable at
room temperature.
[E.2.0] and [ n . b ] systems with sufficiently small n are
also Bredt compounds. However, they are more flexible
than bridged methylenecyclopropanes ( 6 7 ) and cyclopropenes ( 7 1 ) having the same number of bridge atoms.
The recent literature contains reports on the isolation
of ( 8 8 ) and ( 8 9 ) from carbene i s o m e r i z a t i ~ n s [ ~ ~the
”],
&-
..
synthesis of (90 as a derivative of a photoisomerization
ant the intermediate occurrence of the compound (91)[45cl; the higher homolog ( 9 2 ) is also
Angew. Chem. internat. Edit.
Vol. I 2 (1973) / N o . 6
The boundary between stable and unstable
compounds is evidently crossed only with C2.2.01 systems
(93).
carbons. However, one cannot deny the heuristic value
of the Bredt rule as a guideline and stimulus for synthetic
and mechanistic studies.
Received: September 22, 1972;
supplemented: January 25, 1973 [A 937 IE]
German version: Angew. Chem. 85, 494 (1973)
Translated by Express Translation Service, London
10. Outlook
[ I ] J . Brrdr, Liebigs Ann. Chem. 437, 1 (1924).
Scheme 1, to which one could add a long series of heterocyclic and polycyclic analogs, shows 28 bicycles containing
a bridgehead double bond and with S g 7 , which may
nowadays be assumed to be Bredt compounds, or for
which at least the reverse has not yet been proved. insofar
as any conclusion can be drawn from rules A to C, these
compounds are arranged from top to bottom and from
right to left in order of increasing ring strain; the majority
have not yet been investigated. One could naturally argue
[2] J . Bredr, J . Hoirbm, and P . Lrry, Ber. Dtsch. Chem. Ges. 35, 1286
[ 1902).
[3] See, r.g.. H . Mrrrwein, Liebigs Ann. Chem. 405, 134 (1914): E .
Mohr, J. Prakt. Chem [2] 98, 326 (1918): L. Ruzicka and H . Trrdrr,
Helv. Chim. Acta 4, 566 (1921).
[4] F. S. Fawcrtt, Chem. Rev. 47, 219 (1950).
[ S ] G . Wirtig' Stereochemie. Akademische Verlagsgesellschaft, Leipzig
1930, pp. 154f: E . L. E l i d : Stereochemie der Kohlenstoffverbindungen.
Verlag Chemie, Weinheim 1966, pp. 361ff.
161 F . Ebrl in K . Freudmberg' Stereochemie. Verlag Franz Deuticke,
Leipzig 1933, pp. 6491:
s = 7
S = 6
[3.2.1]
cf. (42)
V
s = 5
Q
s = 4
[i.i.o]
Scheme 1
about whether the hypothetical bicycle[ 1.l.O]but-1-ene,
which stands at the end of the series, should really be
classified as a Bredt compound, since its immense ring
strain must be due much more to deformation of the
(J skeleton than to twisting strains on the unsaturated
centers. One can see here the entanglement of our problem
with other aspects of the chemistry of ring-strained hydro472
[7] V Prelog, L. Ruzicka, P. Barman, and L. Frrnkiel, Helv. Chim.
Acta 31, 92 (1948): V Prrlog, P. Barman, and M . Zimmrrmann, ibid.
32, 1284 (1949); see also V. Prrlog, J. Chem. Sac. 1950. 420.
[8] H . Mrerwrin and W. Schurmann, Liebigs Ann. Chem. 398, 196 (1913);
H . Mrrrwrin, F. K i d , G . K l o s g m , and E . Schoch, J. Prakt. Chem. [2]
104, 161 (1922); ! t Prrlog and R . Sriwrrth, Ber. Dtsch. Chem. Ges.
74, 1644 (1941).
[9] J . P. Ferris and N . C . Millrr, J. Amer. Chem. SOC.85, 1325 (1963).
Anguw. Chem. internal. Edit. J Vol. 12 (1973)
1 NO. 6
[lo] F. H. Westheimer and W A. Jones, J. Amer. Chem. SOC.63, 3283
(1941); K . J . Pederson, J . Phys. Chem. 38, 559 (1934).
[ I t ] J . P. Schaefer and J . C. Lark, J . Org. Chem. 30, 1337 (1965).
[I21 J . A. Marshall and H Faubl, J. Amer. Chem. SOC.89, 5965 (1967);
92, 948 ( 1970).
[ 131 a) J . R. Wiseman, J . Arner. Chem. SOC.89,5966 ( I 967); J . R. Wiseman
and W A . Pletcher, ibid. 92, 956 (1970); b) J . R. Wiseman, H.-F. Chan,
and C. J . Ahola, &id. 91, 2812 (1969); c) J . A. Chong and J . R. Wiseman,
ibid. 94, 8627 (1972).
[I41 M . Toda, I! Hirata, and S. Yamamura, Chem. Commun. 1970, 1597.
[I51 A. C. Cope and M . E. Synerholm, J . Amer. Chem. SOC. 72, 5228
( 1950).
[I61 W R. Hatchard and A. K . Schneider, J . Amer. Chem. Soc. 79,
6261 (1957).
[I71 S . Beckmann and 0.S. Ling, Chem. Ber. 94, 1899 (1961).
[I81 J . P. Ferris and N. C. Miller, J. Amer. Chem. SOC.88, 3522 (1966).
1191 P. C . Guha, Ber. Dtsch. Chem. Ges. 72, 1359 (1939).
[20] G. L. Buchanan, N . B. Kean, and R. Taylor, Chem. Commun. 1972,
201; see also G. L. Buchanan and G. Jamieson, Tetrahedron 28, 1123,
I I29 (1972).
[21] C. A. Grob, Angew. Chem. 77,459 (1965); Angew. Chem. internat.
Edit. 4, 441 (1965); C . A. Grob and A. Sieber, Helv. Chim. Acta 50,
2531 (1967).
[22] a) J . 0.Reed and W Lwowski, J . Org. Chem. 36, 2864 (1971);
b) M. Toda, H. Niwa, K . lenaga, and Y. Hirata, Tetrahedron Lett. 1972,
335.
[23] a) S. F . Campbell, R. Stephens, and J . C . Tatlow, Tetrahedron 21,
2997 (1965); S. F. Campbell, J . M . Leach, J . C . Tatlow, and K . N .
Wood, J. Fluorine Chem. I , 103 (1971); b) S. F. Campbell, J . M . Leach,
R. Stephens, and J . C. Tatlow, Tetrahedron Lett. 1967,4269.
[24] a) R. Keese and E.-P. Krebs, Angew. Chem. 83, 254 (1971); Angew.
Chem. internat. Edit. 10, 262 (1971); b) Angew. Chem. 84, 540 (1972);
Angew. Chem. internat. Edit. I I , 518 (1972).
[25] H. H. Grootueld, C. Blomberg, and F . Bickelhaupt, Tetrahedron
Lett. 1971, 1999.
1261 W L. Mock, Tetrahedron Lett. 1972,475.
[27] K . Ziegler and H. Wilms, Liebigs Ann. Chem. 567, 1 (1950); A.
C. Cope, R. A. Pike, and C . F. Spencer, J . Amer. Chem. Soc. 75, 3212
(1953).
[28] a) N . L. Allinger, J . Amer. Chem. SOC. 80, 1953 (1958); b) N.
L. Allinger and J . 7: Sprague, ibid. 94, 5734 (1972).
[29] P . M. Lesko and R. B. Turner, J . Amer. Chem. SOC.90,6888 (1968).
[30] E. J . Corey, F. A. Carey, and R. A. E . Winter, J . Amer. Chem.
SOC.87, 934 (1965).
1311 N . M . Weinshenker and F. D. Greene, J . Amer. Chem. Soc. 90,
506 ( 1968).
[32] D. E. Appleyuist, R. Searle, M.-D. Steinhardt, E. C. Friedrich, and
R. L. Little, J . Org. Chem. 30, 2126 (1965).
Angew. Chem. internat. Edit.
Vol. 12 (1973)
/
No. 6
[33] A. Krebs and H . Kimling, Angew. Chern. 83, 540 (1971); Angew
Chem. internat. Edit. 10, 509 (1971).
[34] G. L. Closs and W A. Biill, J . Amer. Chem. SOC.85, 3904 (1963).
G. L. Closs, W A. Biill, H. Heyn, and V. Dev, &id. 90, 173 (1968).
[35] W Kirmse and K . H . Pook, Angew. Chem. 78, 603 (1966): Angew.
Chem. internat. Edit. 5 , 594 (1966).
[36] K . B. Wiberg, G. J . Burgmaier, and P. Warner, J. Amer. Chem.
SOC.93, 246 ( I97 1).
[37] W R. Moore and T M . Orretich, Tetrahedron Lett. 1967, 3205.
[38] W R. Moore and R. C. Bertelson, J . Org. Chem. 27, 4182 (1962);
L. Skattebel, Acta Chem. Scand. 17, 1683 (1963).
[39] G. Witrig, H. L. Dorsch, and J . Meske-Schiiller, Liebigs Ann. Chem.
711, 55 (1968); G. Wifrig and J . Meske-Schiiller, ibid. 711. 76 (1968);
W R. Moore and W R. Moser, J. Amer. Chem. SOC.92, 5469 (1970);
J. Org. Chem. 35, 908 (1970).
[40] J . P. Visser and J . E . Ramakes, Chem. Commun. 1972, 178.
[41] a) C . L. Osborn, T C . Shields, B. A. Shoulders, J . F . Krause, H.
V Corter, and P. D. Gardner, J . Amer. Chem. SOC. 87, 3158 (1965);
b) W E. Billups, K. H . Leauell, W Y. Chow, and E . S. Lewis, ibid.
94, I770 ( 1972).
[42] 7: C . Shields, B. A. Shoulders, J . F. Krause, C. L. Osborn, and
P. D. Gardner, J . Amer. Chem. Soc. 87, 3026 (1965); T C . Shields and
P. D. Gardner, ibid. 89. 5425 (1967); W E. Billups, A. 3 . Blakeney, and
W Y. Chow, Chem. Commun. 1971, 1461; A. V. Tarakanova, Yu. K .
Grishin, A. G . Vashakidze, E. M. Mil'uitskaya, and A. F . Plath, J . Org.
Chem. USSR 8, 1655 (1972).
[43] 7: Mitsuhashi and W M . Jones, J . Amer. Chem. Soc.94,677 (1972);
G. G. Vander Stouw, A. R. Kraska, and H . Shechter, ibid. 94, 1655 (1972),
and further.literature cited therein.
[44] a) M. Baumann, Dissertation, Technische Universitat Hannover
1972; b) G. Kobrich and M . Baumann, Angew. Chem. 84, 62 (1972);
Angew. Chern. internat. Edit. I I , 52 (1972); c) G. Kiibrich and H. Heinemann, Chem. Commun. 1969,493; d) P . Entmayr, Diplomarbeit, Technische Universitat Hannover 1972; e) R. H. Fischer, unpublished experiments, Heidelberg 1968/69; f )H . Dietrich and G. Kiibrich, to be published.
1451 a) R. A. Moss and J . R. Whittle, Chem. Cornmun. 1969, 341;
R. A. Moss, U.-H. Dolling, and J . R. Whittle, Tetrahedron Lett. 1971,
931; b) R. L. Cargill and A. B. Sears, ibid. 1972, 3555; c) R. Breslow,
W Washburn, and R. G. Bergman, J . Amer. Chem. SOC.91, 196 (1969);
N . L. Bauld, C. E. Dahl, and Y. S. Rim, ibid. 91, 2787 (1969); d) L.
Skattebel and S. Solomon, ibid. 87, 4506 (1965).
1461 a) P. Warner, R. LaRose, C. Lee, and J . C. Clardy, J . Amer. Chem.
SOC.94, 7607 (1972); b) C . B. Reese and M. R. D. Steblrs, J . C. S.
Chem. Commun. 1972, 1231.
[47] a) D. Grant, M. A. McKerury, J . J . Rooney, N. G. Samman, and
G. Step, J. C. S. Chem. Commun. 1972, 1186; b) D.Lenoir, Tetrahedron
Lett. 1972, 4049.
[48] a ) K . W Turnbull, S. J. Could, and D. Arigoni, J. C. S. Chem.
Commun. 1972, 597; b) D. H. Bowen and J . MacMillan, Tetrahedron
Lett. 1972, 41 I I .
473
Документ
Категория
Без категории
Просмотров
1
Размер файла
869 Кб
Теги
compounds, rules, bred
1/--страниц
Пожаловаться на содержимое документа