close

Вход

Забыли?

вход по аккаунту

?

Generation of Carbenes by Thermal Cycloelimination.

код для вставкиСкачать
ANGEWANDTE CHEMIE
VOLUME 1 0 . NUMBER 8
AUGUST 1971
PAGES 529-590
Generation of Carbenes by Thermal Cycloelimination
By Reinhard W. Hoffmannr*]
The growth of interest in carbenes has led to a need for methods that allow their generation in
unimolecular decomposition processes, possible also in the gas phase, without troublesome
by-products. Besides the decomposition of diazo compounds, thermal cycloeliminations from
carbocycles and heterocycles with odd numbers of ring atoms have been found to be suitable for
this purpose. The problems, structural requirements, and scope of such cycloelimination reactions
are discussed in the present progress report. The advantages of carbene generation by thermal
cycloelimination are demonstrated with reference to dialkoxycarbenes, which are very difficult to
obtain in any other way.
1. Introduction
Cycloelimination is the cleavage of a carbocycle or
heterocycle into two or more, usually independent fragments. Cycloelimination is thus defined as the reverse of a
cycloaddition. It can be classified in the same way as the
latter"], and is also subject to the selection rules''] imposed
by the conservation of orbital symmetry.
Cycloeliminations from a ring having an even number of
members can lead to fragments that all contain an even
number of former ring atoms ; the retro-Diels-Alder
reaction serves as an example of a [6+4+2] cycloelimination. Since even fragments, such as r C = C < , NGN,
I
I
>C=C---C=C<,
have saturated valences and are therefore of relatively low energy content, a [6+4+2] cycloelimination is normally favored over a [6-3+3] cycloelirninatiod3], which leads to odd fragments.
[*] Prof. Dr. R. W. Hoffmann
Institut fur Organische Chemie der Universitat
355 Marburg, Lahnberge (Germany)
Angew. Chem. internal. Edit. Vol. 10 (1971) 1 No. 8
Odd fragments generallyI4] do not have their valences
saturated and are therefore reactive intermediates such as
1,3-dipoles, carbenes, or nitrenes. Cycloelimination from
rings with odd numbers of ring atoms must inevitably
yield such odd fragments.
Cycloeliminations lead in a coupled reaction to two or
more fragment^'^]. If one of the fragments is a highenergy, reactive intermediate, thermal cycloelimination
can still be achieved under relatively mild conditions
provided that the other fragment is thermodynamically
very stable, as in the case of molecular nitrogen or benzene.
In cycloeliminations of this type, the exothermic formation
of the stable fragment makes the endothermic formation of
the reactive intermediate possible.
Cycloeliminations may be thermally or photochemically
initiated, and d o not require the presence of other reagents.
It follows from this and from the reasons given earlier that
cycloeliminations provide an ideal method of producing
bifunctional reactive intermediates in the absence of
interfering reactants. Hence, carbenes and 1,3-dipoles, as
well as torsionally strained alkenes, angularly strained
cycloalkynes, and dehydroaromatic compounds, can be
generated, to study their intramolecular and intermolecular
stabilization reactions.
The present report summarizes what we have learned so
far about the production of carbenes by thermal cycloeliminations, while the progress report that follows[fi1
presents a survey of photochemical cycloeliminations.
529
2. Cycloeliminations from Norbornadienes
to phenylacetylene, which led not to ( 3 c ) but to (4)C'I.
Reinvestigation["*
showed that (3c) can be isolated
if the reaction is carried out under mild conditions. Only at
1~--15O"C does the substance decompose with formation
of ( 4 ) and [as a further reaction product of dimethoxycarbene (so)] tetramethoxyethylene (see Scheme 1).
The thermal [5+4+ I]
cycloeliminationof carbon monoxide from norbornadienone ( 1 ) proceeds SO readily that (I)
has so far been intercepted only as a reactive intermediate"],
0
Cleavage of the bridge to give the esters (6) and (7)
competes with this thermal cycloelimination of carbene
(5a).
P-Chloroalkyl esters are similarly formed on therrnolysis
of ring
of the cyclic acetals (3j)--(3n ), The
but has not been isolated"'. The cycloelimination of
carbenes, closely related to carbon monoxide in structure,
CH,O,
,OCH,
C=C,
CH30'
OCH,
(54
(34
should therefore also be possible under mild conditions.
This is demonstrated by the Diels-Alder addition of (2)
c1
COOCH,
(61
(7)
Scheme 1.
CH,O, ,OCH,
cleavage and cycloelimination is typical of the thermolysis
of many 1,2,3,4-tetrachloro-7,7-dialkoxynorbornadienes,
as shown in Table 1.
Closer investigation of the thermal decomposition of (3c)
showed["' that it is a first order reaction and that the rate
of cycloelimination [path (a)] is practically independent
of the polarity of the solvent (kCHaCN/kcyolohexane~4),
whereas the ring cleavage [path (b)] is strongly favored
~400).The cornby polar solvents (kCH3CN/kcyc,ohexane
c1
A
c1
c1
Table 1. Product distribution on thermal decomposition of (3) (cf. Scheme 1).
2R
a
b
c
d
e
f
9
h
i
j
k
I
m
n
2CH,
2CH3
2CH,
2CH,
2CH,
2CH3
2CH3
2CH,
2C,H5CH,
-CH2-CH2-CHz-CH,-CH2--CHz-CH(CH,)-CH(CH3)-CH(CH3)-CH(CH3)-
R'
R2
T
("C)
85
130
100-150
160
140
140
140
H
H
H
Br
H
C6H5
H
CH3
CH,OAc
CH,OAc
CH(CH,)OAc
CH,OAc
SnWH,),
WCH,),
-CO-O-CO[b]
H
C,H,
H
H
H
CH,
H
C6H5
H
H
H
C6H5
190
160-195
120
160
130-170
Yield
Ref
Path (b)
Path (a)
(O<)
50
[121
27
72
see Table 2
[101
38
84
60
-
70
?
?
?
40-60
la1
[a]
94
100
95
105
72
140-150
80
[a1
3:l
[Ill
[13,141
141
[14,151
[I61
[a]
1111
2
[121
-
?
28
10
[111
[Ill
[121
[121
[a] Products of this reaction path were only detected qualitatively.
[b] The free dicarboxylic acid was used for the reaction.
530
Angew. Chem. internat. Edit. 1 Vol. 10 (1971)
/ No. 8
petition between paths (a) and (b) can therefore be controlled by suitable choice of solvent (cf. Table 2), the
cycloelimination predominating in apolar media.
Table 2. Ratio of the reaction paths (a):(b)(cf. Scheme 1) in the thermolysis of (3c) as a function of the solvent 110, 111.
Solvent
Cyclohexane
CCI,
Pure
1,2-Dimethoxyethane
CH,OH
Pyridine
Acetonitrile
12.6:l
1.7:l
1.4:l
0.54:l
0.31 :1
As expected, the cycloelimination of dimethoxycarbene
(5a) becomes increasingly easy with increasing stability
of the resulting aromatic compound. Thus the temperatures
required for cycloelimination from ( I O ) or (11) are
about 100°C higher than thoserequired for cycloelimination
from (3a)[10*2'l,since with (10) or ( I I ) the gain in
aromatization energy is less than on formation of ( 4 ) .
Similarly, difluorocarbene was liberated from ( I 2 ) in
good yields only at temperatures above 450°C[231.Thus it
must be stabilized to a smaller degree than dimethoxycarbene ( 5 a ) .
0.65: 1
0.41 :1
2.1. Mechanism of the Cycloelimination
0.14:l
Assuming that formation of (6) begins with rate-determining cleavage of ( 3 c ) to form the zwitterion (8), it is
understandable that the process is favored by polar
solvents. (8) corresponds to the intermediate of a nucleophilic aromatic substitution, and should therefore readily
lose chloride via (9) to give (6)[17]. In the pyrolysis of
1,2,3,4-tetrachloro-7,7-dimethoxynorbornadiene( 3 a ) in
ethanol, some ethyl 2,3,4-trichlorobenzoate, evidently
formed from a cation analogous to (9), was isolated.
This product could not have been formed by transesterification of the initially formed methyl ester, since this
reaction is too slow under the applied conditions['2].
The cycloelimination of a carbene from a norbornadiene
may proceed in one step as a symmetry-allowed cheletropic
reaction. However, a non-concerted path is nevertheless
possible. This becomes clear when one considers the
cycloeliminations from (3) in the light of the thermal
behavior of other norbornadiene derivatives (see Scheme 2).
R,
R
'
R,R'
(b), R
=
(c). R =
(d), R
=
H, R'
=
0-Alkyl
H,R' = O@
R' = OCH3
Scheme 2.
The ring cleavage [path (b)] is favored not only by polar
solvents but also by substituents on the ring that stabilize a
negative charge. Thus path (b) is considerably more
prominent in the thermolysis of ( 3 b ) than in that of (3a).
Finally, the reaction of (2) with esters of acetylenedicarboxylicacidleads [via the adduct (3),R'=R2 =COOCH,]
only to ring-opened products['* (Similar observations
are reported in ['',221.)
OC OC H,
OCH3
&x
COOCH3
COOCH3
(12). X = C1 or F
x x
Angew. Chem. internal. Edit. / Vol. 10 (1971)
/ No. 8
Thus norbornadiene (13a) isomerizes when heated to give
cycloheptatriene (16a)1241.Similar reactions are found for
7-alkoxynorbornadienes (I3 b) and 7-phenylnorbornadienelZ5],as well as for norbornadien-7-olate (I ~ C ) ' ~ ~ " ' .
The isomerization ( I S ) -+(16) has been formulated as
proceeding via ( I 4 ) and (15) [25*261. A comparison
of the temperatures required for the rearrangement, i.e.
450°C for (13a), 270°C for bornadiene["], 170°C for
(13b)[251,
and 25°C for ( I ~ C ) [ ~showed
~ " ~ ,that the ratedetermining step (13) + (14) is facilitated by radicalstabilizing substituents on C-7. ( I 3 d ) should therefore
isomerize particularly readily; in this case, however,
cycloelimination of dimethoxycarbene (Sa) was observed
on heating at 150°C1281.From the above, it is therefore
deduced that the same step (c) is involved and is ratedetermining in the cycloelimination as well as in the
isomerization. This readily explains why the cycloelimination of (Sa) from (30) proceeds at a temperature about
70°C lower than that required in the case of (13d), since
the four chlorine substituents in ( 3 a ) should considerably
facilitate the formation of the intermediate (14).
According to this view, it is not at the stage (13) but
at (14) that the reaction path branches to give either
cycloelimination or isomerization to (I6). Since the ring
closure of (14) to form norcaradiene (15) [path (d)]
should be exothermic, the cleavage of (14) [path (e)]
531
can compete with path (d) only if it also is exothermic
[unless the temperatures chosen are so high that the
equilibrium between (14) and (16) is established (cf.
Section 3)]. Of course, exothermicity is conveyed to path
(e) not only by the stability of the aromatic fragment
formed but also by that of the carbene.
c1
R
c1
c1
Evidence indicating that the cycloelimination of dialkoxycarbene from (13d) proceeds via (14) as an intermediate can be obtained from intramolecular competition
experiments in which an additional partitioning of the
reaction at (14) is introduced (Scheme 3). The intermediate
( 1 8 a ) r ( 1 4 d ) can thus in principle undergo an irreversible
free-radical elimination to form (19a) [path (01, but this
is only of minor importance compared to cycloelimination
in the gas-phase pyrolysis of (17a)[28-301.
c1
R
R'
ClCH2
H
H
H
H
COOCH3
ClCHz
ClCHz
CH20COCH3
COOCZH,
C6H5
COOCH3
(16d), a process that should give the same products as the
thermolysis of (13d). (16d) does in fact decompose into
dimethoxycarbene and up to 54% of benzene at 130 to
3500~136.371,
The pyrolysis of (16d) also leads to products[381 that
are not formed on thermolysis of (13d). (I6d) cannot
therefore be an intermediate in the pyrolysis of (13d),
as was once suggestedr25!
It is quite possible, however, that the equilibrium between
the compounds of the types (14) and (16) is established
in the cycloelimination of an unstabilized carbene, e.g.
of 2-cyclopenten-I-ylidene from (23) at 350°C[391.
Scheme 3.
Elimination of a radical R or R ' from a dialkoxyalkyl
radical such as (18) should become more facile with
increasing radical stability of R or R1r311.Thus in the
pyrolysis of (17b), the formation of (196) and the cycloelimination of the carbene were equally fast[z9,321,
whereas
in the case of (17c) the formation of (19c) predominated
over the cycloelimination. Elimination of the ally1 radical
thus proceeds at the expense of carbene formation. This is in
accordance with Scheme 3[331,provided the rate of step
(e) does not vary much for the substrates (18a)-(18c).
Quadricyclane derivatives (24) can also be used for
cycloeliminations of this type. As can be seen from Table 3,
this requires higher temperatures, and it can therefore be
assumed that (24) does not decompose directly to products but undergoes fast cycloelimination only after
symmetry-forbidden isomerization to (25). (25 e ) has in
fact been isolated upon careful pyrolysis of (24e) [401.
Finally, other irreversible reactions starting with (14) can
completely suppress cycloelimination,as occurs when (20)
is heated above 220 0C[11,341.
In this case rearrangement
via (21) to (22), without elimination of dichlorocarbene,
is observed.
The hypothesis that the cycloelimination of dimethoxycarbene (5a) from (13d) proceeds via (14d) can also
be confirmed by other experiments, since the steps between
(14) and (16) in Scheme 2 are very probably reversible,
as has been shown for a derivative with R = R 1 =CH3r351.
(14d) should thus be accessible independently by heating
R'
Y
R2
R'
R2
Y
Table 3. Cycloeliminations from quadricyclane derivatives ( 2 4 ) .
Substrate
R'
(a)
(b)
(c)
(d)
(e)
(24)
(261
R2
OCH,
OCH,
-O-CHz-CH,-O-O-CHz-CH2-O-S-CH,-CH2-S-0-C(CHJ-
.
R'
T
("Cj
COOCH,
COOCH,
H
H
COOCH,
200
125
200
200
260
further products
ref
[41]
[41 a1
95
(CH,O),C=C(OCH,),
CO,, CH,=CHz
CO,, CH2=CH,
75
60
csz
1281
[281
(CH,j,C=C=O
~401
(7%
tal
100
[a] Detected qualitatively.
532
Angew. Chem. internat. Edit. / Vol. 10 (1971)
/ No. 8
A carbene functionality at the one-carbon bridge of
norbornadiene (or quadricyclane) is sufficiently rich in
energy to allow the elimination of the bridge as atomic
carbon[" bl.
In the rearrangement of ( 5 b ) , the migration of the allyl
residue to give (28) competes with that of the methyl
group to give (29). The allyl residue migrates only about
0
0
It
II
CHz=CH-CHz-C-OCH3
(28)
0
N - N - T OS
CH3-C-O-CH,-CH=CH,
(29)
twice as rapidly as the methyl, indicating that free-radical
character is not yet very pronounced in the transition
state, i.e. that the transition state of the rearrangement
occurs early on the reaction coordinate. Moreover,
deuterium labeling suggested that in the migration of the
allyl residue, the allyl moiety separates completely from
the remainder of the molecule, as is necessary for a freeradical rearrangement. Thus an equally conceivable [2,3]sigmatropic process was not observed at the high temperature of 240°C[321.
12%
2.2. The Carbenes Liberated
In most of the examples investigated so far, a dialkoxycarbene was liberated by cycloelimination from a 7,7disubstituted norbornadiene. If the cycloelimination leads
to a 1,3-dioxolan-2-ylidene(2-carbena-I,3-dioxolane) (27),
It is not yet certain whether dialkoxycarbenes undergo
this immediately undergoes further cycl~elimination[~~~
a different type of thermal decomposition into alkoxyl
to give carbon dioxide and an
and acyl radicals in the condensed p h a ~ e [ ~ ~In
~ ~any
'!
case, 60-70% of methyl orthoformate (30) is formed
+ coz
on pyrolysis of (3c ) or ( 1 3 d ) in the presence of hydrogen
The formation of this ester could proceed
(27)
as shown in Scheme 4. The propensity of ( 5 a ) to add
methanol is confirmed independently by the fact that ( 5 a )
In the case of dimethoxycarbene (Sa), the nature of the
can be intercepted in good yields as (30) in the pyrolysis
further reaction products is strongly dependent on the
of
( 3 c ) in methanol1"] (cf. also[361).
manner in which the cycloelimination is carried out. If (3c)
XI:-
is pyrolyzed rapidly in the condensed phase, corresponding to a high stationary concentration of ( 5 a ) , the carbene
is almost completely dimerized to tetramethoxyethyleneti0.l1! If on the other hand (3c) is pyrolyzed at a
high dilution in the gas phase, ( 5 a ) isomerizes to form
methyl acetate[''.
431. This rearrangement probably
proceeds via free radicals in which formation of methyl
acetate competes with further fragmentation into carbon
'9
RO-E-OR'
(5)
+
R-
CH3O-E-OCH3
\1 I
RH -R.
HC(OCHd3
(sai
D
- C - O R ~ -+ co2 i
Y
(a), R
(hi. R
(C), R
9
R-C-OR'
~
1
-
R' = CH,
Allyl, R' = CH3
= R' = Ally1
=
=
dioxide and ethaneLz8! This fragmentation is favored by
residues R that form stable free radicals. Thus the fragmentation of (5 b) predominates over the rearrangement
at 24OoC,and (5c) gives only carbon dioxide and biallyIr3'].
CH30H
(30)
"9
+
Q
C H 3 0 i eC-CH3
Scheme 4.
Dialkoxycarbenes have also been intercepted in many
other reactions that leave no doubt as to their formation
in the cycloelimination from 7,7-dialkoxynorbornadienes.
Thus the pyrolysis of ( 3 c ) in the presence of air gave
dimethyl carbonate as an autoxidation product of
(5aj[L0.
I '1, while U,U-dimethyl thiocarbonate was formed
in the presence of sulfur["].A [I+2] cycloaddition of (5a)
to electrophilic olefins has not yet been a ~ h i e v e d ~ " ~ ~ ~ ~ ,
but as one would expect from its nucleophilic character,
( 5 a ) adds to electrophilic multiple bond systems to
form 1,3-dipole~[~~],
the further reaction products of
which have been isolated1461
(Scheme 5). The carbene ( 5 a )
has been acylated in a similar manner.
RO
(5a) + C6H&0C1
R
=
CHs
Angew. Chem. internal. Edit. 1 Vol. 10 (1971) N o . 8
-
RO,
FI
??
+
R-C1 i RO-C-C-C,H,
Scheme 5.
533
The cycloelimination from norbornadiene derivatives can
be used not only for the liberation of dialkoxycarbenes,but
and
also for the production of (CH,)2Si[481,(C6H5)2Sir491,
(CH,),GeLSol,as well as of C6H5P[511and C,H,AS[~~],
the stabilization reactions of which have not yet been
investigated.
3. Cycloelimination from Norcaradienes
The thermal cleavage of 7,7-dimethoxycycloheptatriene
(16d) into dimethoxycarbene (5a) and benzene (Section
I
G
2.1) was interpreted as acycloelimination from a 7,7-dimethoxynorcaradiene (15d). Cycloeliminations that probably
(a), R = H
proceed via norcaradienes can also be achieved from
(b). R = F
other cycloheptatriene derivatives. Thus dimethylamino(c). R = C1
carbene is eliminated, though only in a yield of about
5 %, from 7-dimethylaminocycloheptatrieneat 110°C[53*541.
eliminations, the same should also be expected of norcaradienes that are present entirely as such. This is in fact
so for (38), which yields dichlorocarbene at 150"C[58*
591,
and for (39), which loses benzoylcarbene at 195"C[601.
A temperature of 470°C is required for the cycloelimination
of unsubstituted methylene from cycI~heptatriene[~~~.
These [3-+2+1] cycloeliminations are a reversal of the
On the other hand, no dicyanocarbene is lost from norwell-known [I +2] cycloaddition of carbenes to aromatic
caradiene (40) since isomerization to phenylmalononitrile
compoundsr561.When heated, cycloheptatrienes normally
(Scheme 6 ) occurs very readilyr6'! However, cycloelimisomerize via norcaradienes and biradicals of the type
ination, even of energy-rich carbenes, from norcaradienes
(14) to form derivatives of toluene135! At the reaction
can be achieved photochemically in medium to good
yie1ds[6. 60 - 621
temperature, which is usually higher than in the pyrolysis
of norbornadienes, the intermediate (14) is stabilized by
migration of a hydrogen atom to form (31).
4. Cycloeliminations from Cyclopropanes
H
R-C-R'
OH - 0
R
.
;
,
R
'
.........
(14)
(31)
Scheme 6.
Cycloelimination of carbenes from norcaradienes can
therefore be expected only if the carbene formed is sufficiently stabilized or if the hydrogen shift is not possible
on structural grounds. The latter situation is found in the
following examples.
When (32) is heated at 250°C the intermediate (34) is
formed uiu the norcaradiene (33) and eliminates methylene" '].
In a cycloelimination from a monocyclic system the gain in
stabilization energy in the even fragment can no longer be
as great as in the examples in the last section. It is therefore
to be expected that [3-+2+1] cycloeliminations of this
type can be achieved thermally only if the carbenes liberated
are stabilized and particularly low in energy.
Authentic[631thermal cycloeliminations of carbenes are in
fact known only for fluorinated cyclopropane derivatives,
the deciding factor in this case probably being the low
energy of difluorocarbene. Thus hexafluorocyclopropane
liberates difluorocarbene in a yield of more than 85% at
250°C[64.651, and the carbene dimerizes to tetrafluoroethylene if it is not intercepted by cycloaddition to other
olefins.
The intermediate (36a) that occurs in the thermolysis of
1,6-methano[lO]annulene ( ~ S U )on
, the other hand, does
not liberate methylene but isomerizes to form ( 3 7 ~ ) ' ~ ~ ~ .
Only when the carbene to be eliminated is also stabilized,
as in the cases of (3Sb) and ( ~ S C ) is
, cycloelimination of
the dihalocarbene observed in yields of more than 90% at
250 and 60°C respectively[58s
591.
A cyclopropane ring can in principle be broken in three
ways. Thus dichlorocarbene, chlorofluorocarbene, and
If the norcaradienes that are present only in equilibrium
difluorocarbene could be liberated from (41). As is shown
with cycloheptatrienes undergo thermal [3+2+ I] cyclo534
Angew. Chem. infernat. Edit. J Vol. 10 (1971) J No.8
by the following examples[65- 671, only the most stable
carbene, difluorocarbene, is liberated in such cycloeliminations at 160--270°C.
16* 31, the question has not yet been answered for the thermal
reactions described above.
The cycloelimination of a carbene from diazirines is
particularly favored by the low energy of molecuIar
nitrogen. Diazirines readily yield carbenes both thermally
and photochemically; a preceding electrocyclic ring cleavage to give a diazoalkane has been detected at least in the
latter case[74*751.
F
The structurally similar compound (42) undergoes a clean
cis-trans isomerization in the same temperature range[681
which is best explained by cleavage to form the biradical
(43). Therefore the cycloeliminations from the other
perhalogenated cyclopropanes probably proceed nonconcertedly via an intermediate corresponding to (43).
(42)
’
(43)
Since the cycloeliminations from diazirines have already
761, we shall merely mention here that
been
these reactions are the only established route to mixed
hetero-substituted carbenes such as CH,O-c-F
and
NC-%-F[771
(and potentially also CH,O--%-Cl,
CH,=CH-c-Cl,
or even CH3-c-O-COCH,[781),
and that they allow the study of the interaction of carbenes
with strong acids such as trifluoromethanesulfonic acid[791
or with oxidizing agents such as Cl,, N,O,, or N0,C1[801.
@
6. General Considerations
Strained cyclopropane ring systems such as bicyclo[l.l.0]butane derivatives undergo rearrangements when treated
with [Rh(CO),Cl],, [C,F,Cu],, Cu,Cl,, or HgBr, at temperatures as low as 0-50°C. This rearrangement is
most readily accounted for in terms of a [ 3 + 2 + I]
cycloelimination to form a carbene-metal
Undoubtedly this process is considerably facilitated by the
ring strain in the starting materials.
CH,
CH,
-
CH3
H-shift
..
*
H“kfCH3
IL1 eta1
5. Cycloeliminations from Three-Membered
Heterocycles
The cycloeliminations of carbenes discussed here have
already been considered theoretically as cheletropic reactions[21.For all processes of this type an orbital symmetryallowed path is available, the “choreography” of which
prescribesalinearmovement ofthegroupsfor the[5+4+ I]
cycloelimination and a nonlinear movement for the
[3-2 I] cycloelimination. These hypotheses cannot yet
be verified, owing to lack of stereochemical criteria.
+
The thermal [3+2 + I] cycloelirnination is not confined
to perhalogenated cyclopropanes. Perhalogenated oxiranes
and aziridines also generate the most stable possible
carbene, and do not split off the hetero atom, as is shown by
the following examples[68- 701.
(44)
( 4 4 ) , which is formed only as an intermediate, probably
decomposes similarly into (50) and dimethyl carbonater7‘I.
In the case of the oxiranes and aziridines, the cycloelimination may be preceded by electrocyclic ring opening to
give a 1,3-dip0le[’~! Whereas this has been verified for the
photochemical cycloelimination of carbenes from oxiranes
Angew. Chem. internat. Edit. J Vol. 10 (1971) 1 No. 8
In retrospect, it can be seen that thermal cycloeliminations
allow the liberation of stabilized carbenes, but not of
reactive carbenes such as methylene. The fact that it is
possible in this way to liberate dialkoxycarbenes whose
generation by other routestE1] is unsatisfactory, or to
produce difluorocarbene under mild conditions and in the
gas phase, is not the real incitement of these investigations.
Instead the reason lies in the mechanistic aspects of these
reactions.
The data discussed in Sections 3 and 4 point to a two-step
course for the [3+2+ I] cycloelimination, and are thus
clearly different from the addition of carbenes to olefins,
which proceeds in one step and nonlinearly[82J.However, it
has already been pointed out in a similar connection that
the cycloaddition and its reversal, i. e. the cycloelimination,
need not follow the same reaction pathcE3].
Many more data are available on linear cheletropic
[5+4 + I]cycloeliminations,which are generally one-step
processes[’] if stereochemistry is used as the criterion. It
is therefore the more surprising that the symmetry-allowed
[5 -4 + I] cycloelimination of dialkoxycarbenes ( 5 ) from
7,7-dialkoxynorbornadienes (17) very probably proceeds
in several steps as shown by the data in Section 2. Since
most known electrocyclic reactions strictly follow the
orbital symmetry rules, it is particularly interesting to
535
search for the reasons why some deviate from the allowed
and hence normally energy-favored path.
It may be characteristic of the present system that the
reversal, i. e. the symmetry-allowed [4 + 1+5] cycloaddition of a carbene to an aromatic compound, is unknown. The reactions of carbenes with aromatic compounds
have always been found instead to result in [2+ 1+3]
cycloaddition to give norcaradiene derivatives[56! Perhaps
the energy surface of the norbornadienes permits a symmetry-allowed, low-energy transition into the energy
surface of the norcaradiene~[~’~,
from which, apart from a
nonlinear cheletropic elimination, only a stepwise cycloelimination is possible.
The thermal behavior of the norbornadienes and norcaradienes could however simply be due to the fact that
two carbon-carbon bonds must be broken simultaneously
in a synchronous cycloelimination, whereas the cleavage
of only one bond requires less energy and leads directly to
a stabilized intermediate (14).
The investigations discussed here were carried out at the
Organic Chemistry Institutes of the University of Heidelberg
and of the Technische Hochschule Darinstadt. They were
made possible by support from the Deutsche Forschungsgemeinschaji and the Verband der Chernischen Industrie, which
we gratefully acknowledge.
Received: July 10,1970 [A 827 IE]
German version: Angew Chem. 83,595 (1971)
Translated by Express Translation Service, London
Publication delayed at author’s request in order to allow simultaneous
publication with the following report
[l] R. Hursgen, Angew. Chem. 80, 329 (1968); Angew. Chem. internat.
Edit. 7, 321 (1968).
[2] R . B. Woodward and R. Hofmann, Angew. Cbem. 81, 797 (1969);
Angew. Chem. internat. Edit. 8, 781 (1969).
[3] An example may be found in R. Grashey and K . Adelsberger,
Angew. Chem. 74, 292 (1962); Angew. Chem. internat. Edit. I , 267
(1962).
[4] Exceptions are, e.g. SO,, CO, and R-N=C.
[ 5 ] B. P. Stork and A. J . Duke: Extrusion Reactions, Pergamon Press,
Oxford 1967.
[6] G. W. Griffin, Angew. Chem. 83,604 (1971);Angew. Chem. internat.
Edit. 10, 537 (1971).
[7] J . M . Landesberg and J . Sieczkowski, J. Amer. Chem. SOC.93,972
(1971).
181 Cf. J . Meinwald and E. G . Miller, Tetrahedron Lett. 1961, 253;
S. Yankeluich and B. Fuchs, ibid. 1967, 4945.
[9] E. T. McBee, W. R. Diueley, and J . E. Burch, J. Amer. Chem. SOC.77,
385 (1955); see also E. T. McBee, J . D.Idol, and C. W. Roberts, ibid. 77,
6674 (1955).
[lo] R. W. Hoffmann and H . Hauser, Tetrahedron 21, 891 (1965).
[ll] D. M . Lemal, E. P. Gosselink, and S. D. McGregor, J. Amer. Chem.
SOC.88, 582 (1966).
[I21 K . Mackenzie, J. Chem. SOC.1964, Suppl. 1, 5710.
[I31 H . Feichtinger and H . Linden, German Pat. I105862 (1961);
Chem. Abstr. 56, 12803 (1962).
[I41 The substrate was decomposed in situ in the synthesis.
[I51 D. Seyferth and A. B. Ecnin, I. Amer. Chem. SOC.89, 1468 (1967).
[IS] A . P. Stefani and L. C. Daniel, J. Miss. Acad. Sci. 13, 82 (1965).
[17] A competing cleavage of dimetboxycarbene from (8) cannot be
completely ruled out, but it would be difficult to explain the solvent
dependence of the product distribution (Table 2) if (8) decomposed
appreciably into ( 5 ) and ( 4 ) .
[I81 H. Feichtinger and H. Linden, German Pat. I087590 (1960).
Chem. Abstr. 55, 16489 (1961).
[I91 J . Diekmann, J. Org. Chem. 28,2880 (1963); R. G. Pews, C . W Roberts, and C . R. Hand, Tetrahedron 26. 1711 11970).
536
[20] Contradictory results are reported by A. P. Stefani and L. C.
Daniel [16].
[21] P. Kniel, Helv. Chim. Acta 46, 492 (1963); 48, 837 (1965); R. E.
Winkler, ibid. 50, 2497 (1967).
[22] T Jaworski and W Polaczkowa, Roczniki Chem. 34, 887 (1960);
Chem. Abstr. 55, 8407 (1961); R. G . Pews, E. B. Nyquist, and F. P.
Corson, I. Org. Chem. 35, 4096 (1970).
[23] E. T McBee, D. K . Smith, and H . E. Unqnade, J. Amer. Chem. SOC.
77, 387 (1955); R. E. Banks, A. C. Harrison. R. N . Haszeldine, and K . G.
Orrell, I. Chem. SOC.C 1967, 1608.
[24] W G. Woods, J. Org. Chem. 23, 110 (1958).
[25] R. K . Lustgarten and H. G. Richey jr., Tetrahedron Lett. 1966,
4655.
[25a] B. Franzus, W C. Baird jr., R. E. Felty, J . C. Smith, and M. L .
Scheinbaum, Tetrahedron Lett. 1971, 295.
1261 W C. Herndon and L. Lowry, J. Amer. Chem. SOC.86,1922 (1964).
[27] M . R. Willcott and C. J . Boriack, 5. Amer. Chem. SOC.90, 3287
(1968).
[28] D. M . Lemal, R. A . Lomld and R. W Harrington, Tetrahedron
Lett. 1965, 2779.
[29] R. W Hofmann and R. Hirsch, Tetrahedron Lett. 1970,4819.
[30] P. G. Gassman, D. H . Aue, and D. S. Patton, J. Amer. Chem. SOC.
90, 7271 (1968).
[31] E. S. Huyser and D. T Wang, J. Org. Chem. 29,2720 (1964).
[32] R. Hirsch. Dissertation Universitat Heidelberg 1970.
[33] If the cycloelimination proceeded in one step [path ic)1 and were
Independent of (18), this result could be explained only if step (c)
were fast in relation to (r) and (g), and if it were reversible. N o evidence
of reversibilitji of step (c) has so far appeared in the literature. On the
contrary, the step corresponding to (c) has been shown to be irreversible
in the thermolysis of bornadiene: M . R. Willcott and C. J . Boriack, J.
Amer. Chem. SOC.93, 2354 (1971).
1341 H M. Moiotsky, US-Pat. 2946817 (1960); Chem. Abstr. 55,
5856 (1961).
[35] J . A. Berson and M . R. Willcort, J. Amer. Chem. SOC.88, 2494
(1966).
[36] R . W Hofmann and J . Schneider, Tetrahedron Lett. 1967, 4347.
[37] Tropone ethylene ketal also decomposes into benzene, ethylene,
and carbon dioxide to an extent of 36% at 140°C; 7: Fukunaga, T.
Mukai, Y Akasaki, and R. Suzuki, Tetrahedron Lett. 1970, 2975.
[38] R. W Hofmann, K. R. Eicken, H . J . Luthardt. and B. Dittrich,
Chem. Ber. 103, 1547 (1970).
1391 R. Criegee, F. Fsrg, H . A . Brune, and D.Schiinieber, Chem. Ber. 97,
3461 (1964).
[40] R. Schuifler, Diplomarbeit, Technische Hochschule Darmstadt
1970.
[41] R. Hirsch, Diplomarbeit, Universitat Heidelberg 1968.
[41a] H . Prinzbach, J . Ricier, and G. Englert, Helv. Chim. Acta 53,
2219 (1970).
[41b] P. B. Sheclin and A . P. WoS Tetrahedron Lett. 1970, 3987;
see also R. A. Moss, U . H. Dolling, and J . R. Whittle. ibid. 1971, 931.
[42] E. J . Corey and R. A. E. Winter, J. Amer. Chem. SOC.85, 2677
(1963); G Crank and F. W Eastwood, Australian J. Chem. 17, 1392
(1964).
[43] R . W Hofmann and C. W n s c h e , Chem. Ber. 100,943 (1967).
[44] Cf. R. J . Crawford and R. Raap, Proc. Chem. SOC.(London) 1963,
370
[45] Decomposition of ( 5 a ) into dimethyl ether and carbon monoxide
has been discussed [13], but has not been verified, since the detection
reactions used would also have responded to CH,CI and to tetramethoxyethylene respectively ; H . Feichtinger, personal communication.
[46] R. W Hofmann, B. Dittrich, and K . Steinbach, unpublished.
[47] For similar reactions with isocyanides, see K . Ley, U . Eholzer,
and R . Nasr, Angew. Chem. 77, 544 (1965); Angew. Chem. internal.
Edit. 4, 519 (1965); E. Winterfeldt, D. Schumann, and H . J . Dillinger,
Chem. Ber. 102, 1656 (1969); J . Goerdeler and D. Wobig, Liebigs Ann.
Chem. 731,120 (1970).
[48] H. Gilman, S . G. Cortis, and W. H . Arwell, I. Amer. Chem. SOC.86,
1596 (1964); R . Maruca, J. Org. Cheni 36. 1676 ( 1 Y i l )
[49] H . Gilman, S . G. Cortis, and W H . Atwell, I. Amer. Chem. SOC.86,
5584 (1964).
[SO] 0. M . Nefedow and M . N . Manakow, Angew. Chem. 78, 1039
(1966); Angew. Chem. internat. Edit. 5 , 1021 (1966); J . G . Zacistoski
and J . J . Zuckerman, J. Amer. Chem. SOC.90, 6612 (1968).
[51] E H . Braye, W Hiibel, and I . Caplier, I. Amer. Chem. SOC 83,
4406 (1961); 1. G. M . Campbell, R. C. Cookson, M . 8. Hocking, and
Angew. Chem. internat. Edit. / Vol. I0 (1971) / N o . 8
A . N . Hughes, J. Chem. SOC.1965,2184; U . Schmidt, I . Boie, C.Osterroht,
R. Schroer, and H . F. Griitzmacher, Chem. Ber. 101, 1381 (1968).
[52] G . Markl und H . Hauptmann, Tetrahedron Lett. 1968, 3257.
[53] A . P. ter Borg, E. Razenberg, and H . Kloosterziel, Rec. Trav.
Chim. Pays-Bas 84, 1230 (1965); 85, 774 (1966).
[54] Cf. also the analogous cycloelimination of C,H,PO and CIP
from a phosphepane derivative: G. Mark1 and H . Schuberr, Tetrahedron
Lett. 1970, 1237; G. Markl and A . Merz, rbid. 1971, 1269.
[ S S ] 0. Nefrdow. N . Nowizkaja, and A. Iwaschenko, Liebigs Ann.
Chem. 707, 217 (1967).
[56] W Kirmse: Carbene. Carbenoide und Carbenanaloge. Verlag
Chemie, Weinheim 1969. p. 57.
[57] H . B. Renfroe, J. Amer Chem. SOC. 90, 2194 (1968); cf. also
V. Boeckelheide, Proc. R. A. Welch Foundation Conferences, XII, 109
(1969).
[ 5 8 ] E. Vogel, Pure Appl. Chem. 20, 237 (1969); Proc. R. A. Welch
Foundation Conferences, XII, 225 (1969).
[59] I/. Rautensrrauch, H. J . Scholl, and E . Vogel, Angew. Chem. 80,
278 (1968); Angew. Chem. internat. Edit. 7, 288 (1968).
1601 U . Heep, Dissertation, Universitat Karlsruhe 1968.
[61] E . Ciganek, J. Amer. Chem. SOC.89, 1458 (1967).
[62] D. B. Richardson, L. R . Durreri, J . M . Martin jr., W E. Putnam,
S . C. Slaymaker, and I . Droretzky, J. Amer. Chem. SOC.87,2763 (1965);
M . Pomerantz and G . W Gruber, ibid. 89, 6798 (1967); J . S. Swenton
and A. J . Kruibsack, ibrd. 91,786 (1969); T Toda, M . Nitta, and T Mukai,
Tetrahedron Lett. 1969,4401 ; H. Diirr and G. Scheppers, Liebigs Ann.
Chem. 734, 141 (1970).
[63] Some thermal cycloeliminations reported in the literature for
cyclopropane derivatives have not been confirmed: R. H. Eastman
and A . V. Winn, J. Amer. Chem. SOC.82, 5908 (1960); W c. E . Doering,
M . R. Willcott 111, and :\I. Juries jr., ibid. 84, 1224 (19021. a n d : A
Schiinberg, A. Mustafa, and N . Latif, ibid. 75, 2267 (1953); G. W Griffin,
personal communication 1969.
[64] B. Atkinson and D. McKeagan, Chem. Commun. 1966, 189.
[65] J . M . Birchall, R . N . Haszeldine, and D. W Roberts, Chem.
Commun. 1967,287.
[66] R. N. Haszeldine and J . G. Speight, Chem. Commun. 1967. 995.
[67] R. A. Mitsch and E . M! Neuoar, J. Phys. Chem. 70, 546 (1966).
[68] P. B. Sargeant, J. Org. Chem. 35,678 (1970).
16Xal 1'. G . Gassman and F. J . Williams, J. Amer. Chem. SOC.92, 7631
(1970); P. G. Gassman, T J . Atkins, and F. J . Williams, ibid. 93, 1812
(1971); P. G. Gassman and F. J . Williams.Tetrahedron Lett. 1971,1409:
P. G. Gassman and F. A. Armour, ;bid 1971, 1431 ; M . Sakai. H . Yamclguuchi, and S . Masamune, Chem. Commun. 1971,486; L . A. Paquetrr,.
R . P. Henzel, and S. E. Wilson, J. Amer. Chem. SOC.93,2335 (1971).
[69] R . A. Mitsch, E. W Neucar, and P. H. Ogden, J. Heterocycl.
Chem. 4, 389 (1967).
[70] W R. Brasen, H. N . Cripps, C. G. Bottomlr): M . W Furlow, and
C.G. Krespan, J. Org. Chem. 30,4188 (1965).
[71] R. W Hoffrnann and J . Schneider, Chem. Ber. 100, 3698 (1967).
[72] P. Brown and R . C Cookson, Proc. Chem. SOC.(London) 1964.
185; W J . Linn, J. Amer. Chem. SOC.87, 3665 (1965); H . W Heitre.
R. Peacy, and A . J . Durbetaki, J. Org. Chem. 31,3924 (1966);R. Huisgrn,
M! Scheer, and H . Huher, J Amer. Chem. SOC.89, 1753 (1967).
[73] T Do-Mrnh, A . M . Ttozzolo, and G. W Grlflk, J. Amer. Chem. SOC
92, 1404 (1970).
[74] H . M . Frey, Advan. Photochem. 4, 225 (1966).
1751 P. H. Ogden and R. A . Mitsch, J. Heterocycl. Chem. 5, 41 (1968).
[76] E. Schmitz, Angew. Chem. 76,197 (1964);Angew. Chem. internat.
Edit. 3, 333 (1964).
[77] R. A. Mitsch, E. W Neucar, R. J Koshar, and D. H. Dyhwg.
J. Heterocycl. Chem. 2, 371 (1965).
[78] W H. Grahum, J. Amer. Chem. SOC.87,4396 (1965).
[79] R. A. Mirsch and J . E. Robertson, J. Heterocycl. Chem. 2. 152
(1965).
[80] R . A. Mitsch, J. Heterocycl. Chem. I , 233 (1964).
[81] R . A . OloJson, S. W Waiinskp, J . P . Murino, and J . L Jrrnow,
J. Amer. Chem. SOC.90,6554 (1968).
[82] R. Hofmann, J. Amer. Chem. SOC.90, 1475 (1968).
[83] R . Huisgen, L. A Feilrr, and G. Binsch, Chem. Ber. l02, 3460
(1969).
Generation of Carbenes by Photochemical Cycloelimination
By Gary W. Griffin'*]
A wide variety of shelfstable substrates undergo photochemical cycloelimination under relatively
mild conditions. Suitable substrates include cyclopropanes, oxiranes, aziridines, and diazirines,
as well as 3H-pyrazoles and IJ-dioxolanes. Carbenes prepared from these substrates as well as
certain cyclic carbonates, sulfites, and phosphoranes behave chemically in a manner similar to
divalent carbon species produced from conventional diazo precursors, namely, they react with
alkenes to give cyclopropanes and undergo insertion into C-H bonds of alkanes. In many cases
the carbenesformed may also be detected by EPR and optical spectroscopic methods.
1. Introduction
It has become increasingly apparent that divalent carbon
intermediates play an important role in the photochemistry
ofa variety of organic substrates. In many cases the reactions
have synthetic utility which may complement conventional
[*I
Prof. Dr. G. W. Griffin
Louisiana State University in New Orleans
Department of Chemistry
Lake Front, New Orleans, Louisiana 70 122 (USA)
Angew. Chem. internat. Edit. 1 Vol. 10 (1971) 1 No. 8
techniques for the generation of carbenes. It is our intent
in this review to summarize those examples of photoinduced cycloeliminations which lead to carbene formation.
In accordance with the precedent established in the related
review""] on thermolytic cycloeliminations we shall employ
the convention suggested by Huisgen" b1 in designating the
mode of cycloelimination.
To the extent that the reactions described herein are
concerted they are subject to orbital symmetry
constraints"'! For example the [3+2 + I]
cycloelimination
reaction of a cyclopropane to give a carbene if entirely
531
Документ
Категория
Без категории
Просмотров
0
Размер файла
799 Кб
Теги
thermal, carbene, generation, cycloelimination
1/--страниц
Пожаловаться на содержимое документа