вход по аккаунту


Catalytic Isomerization of Methylated 1 5-Cyclooctadienes.

код для вставкиСкачать
pentalene 2 is estimated by comparing the chemical shift
of the ring proton with that of the C H = N protons of 4
(6 = 7.23).[9'The extension of the conjugation by introduction of double bonds in 4 should induce a downfield shift
of the proton signal. Assuming that the chemical shift difference is the same as that (A6= 0.82 ppm) for the olefinic
protons on going from cyclopentene (6= 5.60) to cyclopentadiene (6=6.42 ppm), the signal for the ring proton of 2
is expected to be located at 6=8.05 ppm. The difference
between the expected and observed value (A6= 0.88 ppm)
can be attributed to the paramagnetic shielding effect in
this diazapentalene. The degree of the shielding is comparable to that in 1,4-dihydropyrazines ( A 6 ~ 0 . 7 5ppm),
which is considered to be a typical antiaromatic molecule."']
plexes, undergoes facile rearrangement in the presence of
catalysts to give bicyclo[3.3.0]oct-2-ene 2. This reaction
may be effected in a particularly selective fashion under
mild conditions using a catalyst prepared from bis(2-ethylhexanoato)nickel and dichloroethylaluminum, as described by Farona et al."] The analogous reaction with substituted 1,5-cyclooctadienes can, depending on the pattern
of substitution, lead to numerous isomers, analysis of
which affords insight into the stereochemical course of the
catalytic isomerization.
Isoprene reacts on diazadieneiron(0) catalysts at room
temperature to give directly 1,6-dimethyl-l,5-cyclooctadiene 3, and a 1 : 1 mixture of isoprene and trans-piperylene selectively affords 1,7-dimethyl-1,5-cyclooctadiene
4.[21Piperylene and 2,3-dimethylbutadiene also react to
give codimers (89%); these consist mainly (70%) of 1,2,4trimethyl-1,5-cyclooctadiene 5 (62% overall yield). All
three cyclooctadienes react on the Farona catalyst to give
unexpected products, which depend on the solvent and
temperature. The topology and stereochemistry of the
products could be elucidated by 2 D INADEQUATE exp e r i m e n t ~ ' and
~ ] other 2D NMR methods.
Received: March 15, 1988 [Z 2664 IE]
German version: Angew. Chem. 100 (1988) 1134
[ I ] a)'B. A. Hess, Jr., L. J. Schaad, C. W. Holyoke, Jr., Tetrahedron31 (1975)
295: b) I. Gutmann, M. Milun, N. Trinajstic, J . Am. Chem. SOC.99
(1977) 1692.
[2] a) K. Hafner, F. Schmidt, Angew. Chem. 85 (1973) 450; Angew. Chem.
Inl. Ed. Engl. 12 (1973) 418; b) H.-J. Gais, K. Hafner, Tetrahedron Lett.
1974, 771; c ) W. Treibs, Nuturwissenschafien 46 (1959) 170; d) S. Hiinig,
H.-C. Steinmetzer, Justus Liebigs Arm. Cbem 1976, 1090; e) J . J. Eisch,
T. Abraham, Tetrahedron Lett. 1976, 1647; 1) F. Closs, R. Gompper, Angew. Chem. 99 (1987) 564: Angew. Chem. I n / . Ed. Engl. 26 (1987) 552.
[3] K. Satake, T. Kumagai, T. Mukai, Chem. Lett. 1984, 2033.
141 a) B. Kitschke, H. J . Linder, Tefruhedron Let!. 1977, 251 1 ; b) P. Bischop,
R. Gleiter, K. Hafner, K. H. Knauer, J . Spanget-Larsen, H. U. Suss,
Chem. Ber. 111 (1978) 932 and Ref. [20] therein.
[5] W. Adam, A. Grimison, G. Rodriguez, Tetrahedron 23 (1967) 2513.
[6] G. Binsch, I. Tamir, J. Am. Chem. Soc. 91 (1969) 2450.
171 a) K. Hafner, H. U. Suss, Angew. Chem. 85 (1973) 626; Angew. Chem.
Int. Ed. Engi. 12 (1973) 575: b) T. Nakajima, Y. Yaguchi, R. Kaeriyama,
Y. Nemoto, BuU. Chem. SOC.Jpn. 37 (1964) 272; c) N. C. Baird, R. M.
West, J. Am. Chem. SOC.93 (1971) 3072.
[S] a) J. H. van Vleck: The Theory OfElectronic and Magnetic Susceptibikies.
Clarendon, Oxford 1932, Chap. 4; b) J. A. Pople, K. G. Untch, J. An!.
Chem. SOC.88 (1966) 4811.
191 K. Satake, T. Kumagai, T. Mukai, Chem. Lett. 1983, 743; The NMR
spectra of 2 and 4 were measured under the same conditions.
[lo] W. Kaim, Angew. Chem. 93 (1981) 620; Angew. Chem. lnt. Ed. Engl. 20
(1981) 599.
Heindirk torn Dieck*
Dedicated to Professor Heinrich Noth
on the occasion of his 60th birthday
1,5-Cyclooctadiene 1 , which is readily accessible by catalytic cyclodimerization of butadiene on nickel com[*] Prof. Dr. H. tom Dieck, DipLChem. M. Mallien, Dr. E. T. K. Haupt
lnstitut fur Anorganische und
Angewandte Chemie der Universitat
Manin-Luther-King-Platz 6, D-2000 Hamburg 13 (FRG)
This work was supported by the Deutsche Forschungsgemeinschaft, the
Fonds der Chemischen Industrie, and the Stiftung Volkswagenwerk.
0 VCH Veriagsgeseiischafl mbH. 0.6940 Weinheim, 1988
Catalytic Isomerization of Methylated
By Michael Mallien, Erhard T.K . Haupt, and
In benzene, 3 reacts at 81°C to give 2,5-dimethylbicyclo[3.3.0]oct-2-ene 6 as the main product (93% after 4 h);
this corresponds to the transformation 1-2. At 69°C in
hexane, on the other hand, 1,4-dimethylbicyclo[2.2.2]oct-2ene 7 is found (63% after 8 h). Its formation cannot be explained alone by C-C bond formation and hydride shift;
C-C bond breaking is also required. Compound 6 does
not isomerize to 7 in refluxing benzene.
Reaction of 4 for a longer time in benzene yields primarily 5,7-dimethylbicyclo[3.3.0]oct-2-ene8. The reaction appears to be analogous to that of 1 to 2 and of 3 to 6.
However, in contrast to 6, the methyl group is not located
in the unsaturated but rather in the saturated five-mem-
0570-0833/88i0808-1062 $ 02.50/0
Angew. Chem. Int. Ed. Engl. 27 (1988) No. 8
bered ring. Reaction of 4 in refluxing hexane, on the other
hand, gives a fully saturated compound as the main product, which was identified as 2,4-dimethyltricyclo[5. 1.0.04~8]octane9. The catalytic formation of such a
highly strained cyclopropane system may at first seem surprising. Reaction times longer than 2 h, however, result in
the slow isomerization of 9 to 8 (50% after 40 h) and the
formation of six further, unidentified products having similar retention times upon gas-chromatographic separation.
Reaction of 5 in refluxing hexane also results in rapid and
exclusive formation of a tricyclic compound, 2,4,8-trimethyltricyclo[5.]octane 10, which, on first sight, resembles 9. The positions of the methyl groups reveal once
again that the route to 10 does not occur only via the formation of C-C bonds but also by bond breaking.
A nickel hydrido complex, which coordinates 1 and 3-5
is presumably the active species involved in the isomerization, since other nickel hydrido systems also allow the
isomerization of 1 to 2.14]The first step, after coordination
of the cyclooctadiene, is a hydrogen transfer from the
nickel center to one of the sp2 carbon atoms. A Ni-C CF
bond to a tertiary C atom is thereby formed (A, A’), since
the products 6-10 all have at least one angular methyl
group. Because bonds to secondary C atoms should be
more favorable, the activation barrier for a product with a
Ni-C,,, 5 bond in the catalytic cycle must be higher and
the yield of the corresponding catalysis product lower.
Formation of a transannular C-C bond results in transformation of A to the bicyclo[3.3.0]octylnickel intermediate
B, from which 2 or 6 are formed by p-elimination. In the
corresponding intermediate involved in the isomerization
of 4, the nickel center would have a methyl group and a
bridgehead H atom as neighbors. If the methyl group is cis
to the nickel atom, then no &elimination to this side is
possible, so that a product containing a double bond to a
bridgehead C atom should be formed. This does not occur,
but instead 9 is obtained. Its formation can easily be explained by an intermediate containing a methyl group and
a nickel center in endo positions, B‘, since y-elimination
from such a species with formation of the cyclopropane
ring of 9 should be sterically facile. However, a n endo Ni
atom can only be attained by opening of the eight-membered ring (e.g., A ‘ L F L B ’ ) . The Ni-H elimination leading to 9 is apparently reversible, as shown by the isomerization of 9 to 8. If “Ni-H” adds to 9 in such a way that
G and not B ’ is formed, a normal 8-elimination can readily occur leading to 8. This mechanism would explain
why-in contrast to 6-compound 8 contains the double
bond in the unsubstituted five-membered ring.
The formation of 7 formally requires, among other
steps, the opening of the C3-C4 bond of 3 to give C. A
second hydride transfer would give the new enyl complex
D. A bicyclo[2.2.2]octylnickel complex can be formed in
one step from this species. Ni-H elimination from this
complex, in turn, gives 7.Evidence for the intermediacy of
C is provided by the appearance of traces of a product
having the retention time of a 1,4-dimethyl-4-vinylcyclohexene. However, a catalytic rearrangement of 1,4-dimethyl-4-vinylcyclohexene itself to 7 cannot be detected.
Finally, the formation of the tricyclic compound 10
from 5 is a complete surprise. The INADEQUATE experiment, in connection with a I3C AFT spectrum
(APT = attached proton test) unambiguously gives, without any presuppositions, the molecular topology. Two of
the methyl groups are found at mutually linked quaternary
C atoms ; the molecule has a three-membered ring containing a quaternary and two tertiary C atoms; the third meAngew. Chem. Inr. Ed. Engl. 27 (1988) No. 8
1.2, R = H
3,6. R = Me
R 2,6
thy1 group is found on a tertiary C atom, the configuration
of which cannot be derived with certainty from the ’HN M R data, and is in proximity to the three-membered
ring. Therefore, this tricyclic structure must have been
formed by (formal!) breaking of the Cl-C8 bond and formation of the bonds C2-C8 as well as CILC5 and Cl-C6
(numbering of 5). The formation of 9 could also have occurred in a sequence of steps analogous to the path 5- 10
instead of the “simple” way via F, B‘.
The formation of 7 and 10 clearly shows that the primary coordination of the active nickel species is capable of
activating different (T bonds of the eight-membered ring.
The formation of methylene-2-vinylcyclopentane, often
observed upon dimerization of butadiene to the usual cyclodimers, can also be explained by opening of the C3-C4
bond of B and Ni-H elimination. The extremely selective
formation of the various types of products as a function of
the pattern of substitution and the temperature, however,
still remains to be explained.
Experimental Procedure
The 13C-NMR assignment for 2 reported by Benn et al. [6a] and by Whitesell
et al. [6b] could be confirmed.
6 , 7: Bis(2-ethylhexanoato)nickel(0.1 mmol) [l] as a 0.1 M solution in hexane
was added to 3 (2.7 g, 20 mmol) in hexane (15 mL) and activated with 2 m L
of a 1 M solution of EtAICI, (2 mmol) in hexane. After refluxing for 9 h, the
reaction was quenched by adding 5 mL of EtOH. 3 had reacted completely,
resulting in the formation of 7 (63%), 6 (7%), and three further products in
7-8% yield. The same reaction in benzene yielded 6 (93%) along with three
further products (together 7%).-‘”C-NMR (CDCI,): 6 : 6 = 15.45 ((29). 25.90
(C7), 28.92 (‘210). 30.80 (C8), 43.05 (C6), 48.25 (C4), 49.07 (C5). 61.14 (CI),
122.71 (C3), 141.80 (C2); the ”C-NMR assignment in [6b], which corrected
that in [6c], was confirmed by 2D INADEQUATE NMR experiments. 7:
6=25.37 (C9, ClO), 33.95 (CI, C4), 35.04 (C5, C6, C7, C8), 138.21 (C2, C3).
Hydrogenation afforded 1,4-dimethylbicyclo[2.2.2]octane, whose ”C-NM R
spectrum is in agreement with the data given in IS].
8, 9 : The precatalyst and activator were added to 4 (2.7 g, 20 mmol) in
15 rnL of hexane as described above. After refluxing for 2 h, 60% of 4 had
reacted to give virtually exclusively 9 . After 4-5 h, 4 had undergone complete reaction. Analysis of the products showed the presence of 9 (63%). 8
(lo%), and unidentifiable products having similar retention times (16, 5 , 5 , 3,
and 2%). After 40 h, the proportion of 8 had attained 50°A and 9 was no
0 VCH Verlagsgesellschaji mbH, 0-6940 Weinheim. 1988
0570-0833/88/0808-1063 $ 02.50/0
longer present. In addition, isomers were formed in yields of 16,5, 5, and 4%
along with further compounds (<3%). The course of reaction is similar, but
faster, in refluxing benzene or heptane.-”C-NMR (CDC13): 8 : 6= 19.10
(CIO), 29.51 (C9), 33.52 (C7), 41.14 (C8), 48.10 (CS), 49.80 (C4), 52.27 (C6),
58 31 (CI), 129.35 (C3), 134.13 (C2). 9 : 6=22.19 (ClO), 24.29 (C6), 25.27
(C9), 26.97 (C7), 31.35 (C2), 34.08 (CI), 35.70 (CS), 47.40 (C5), 50.99 (C4),
57.06 (C3).
10: 5 was prepared according to the procedure in [2] by reaction of 2,3dimethylbutadiene and frans-piperylene (30 mmol each) with (dad)*Fe/AIEt,
I mmol Fe/5 mmol Al) at
(dad = 1,4-dicyclohexyl-1,4-diaza-l,3-butadiene;
room temperature for 60 h (after complete conversion: 63% 5,22% 1,Z-dimethyl-4-propenylcyclohexenes(DMPC), 9% piperylene codimer, and 6% of an
unidentified codimer). Reaction of the mixture of 5 and DMPC with the
nickel catalyst in refluxing hexane for 2 h resulted in exclusive and complete
conversion to 10 together with unreacted DMPC. After 3 additional hours of
refluxing, another isomerization occurs which has not yet been identified.
10: I3C-NMR (CDCI,): 6= 16.59 (Cll), 22.50 (C9), 22.98 (C6), 23.24 (CIO),
30.18 (CZ), 33.76 (C7), 38.72 (CS), 40.79 (CI), 47.21 (C5), 51.66 (C4), 56.76
ping reactions.l7I The deformation of polycyclic arenes, on
the other hand, has been little investigated.[’]
We have already described the synthesis of [2](2,6)naphthalino[2]paracyclophane 2 from the dithia[3.3]phanes 4
and 5 , which were ring-contracted by photolytic sulfur extrusion and pyrolytic sulfur dioxide extrusion, respectiveIY.[~] Since, formally, 2 can be viewed as the product in
Received: March 7, 1988;
revised version: May 17, 1988 [Z 2648 IE]
German version: Angew. Chem. 100 (1988) 1091
[I] N. A. Maly, H. R. Menapace, M. F. Farona, J. Caral. 29 (1973) 182.
[2] H. tom Dieck, J. Dietrich, Angew. Chem. 97 (1985) 795; Angew. Chem.
Int. Ed Engl. 24 (1985) 781.
[3] J. Buddrus, H. Bauer, Angew. Chem. 99 (1987) 642: Angew. Chem. Inr. Ed.
Engl. 26 (1987) 625.
[4] W. Keim, New J. Chem. I 1 (1987) 531.
[5] W. Adcock, A. N. Abeywickrema, V. Sankarlyer, G. B. Kok, Magn. Resun. Chem. 24 (1986) 213.
[6] a) R. Benn, H. Butenschon, R. Mynott, W. Wisniewski, Magn. Resun.
Chem. 25 (1987) 653; b) J. K. Whitesell, R. S . Matthews, J. Org. Chem. 42
(1977) 3878; c) J. K. Whitesell, R. S . Matthews, P. A. Solomon, Terrahedrun Letf 1976. 1549.
Strongly Bent Naphthalene Units in
6, X= 5
x = so*
x = so
which the two ends of the longer 2,6-dimethylnaphthalene
are linked with the ends of the shorter p-xylene, a “bow”
(naphthalene) and a “bow string” (benzene) must be
formed. Indeed, an X-ray structure analysis of 2 (Fig. 1)
reveals an extremely bent naphthalene moiety bridged by a
nearly planar benzene
By Norman E. Blank, Matthias W. Haenel,* Carl Kruger,
Yi-Hung Tsay, and Heike Wientges
In the series of [nlparacyclophanes 1, n=4-8, the decreasing length of the aliphatic bridge increasingly forces
the benzene ring to adopt a deformed boatlike structure.
The effects of this deformation on the spectroscopic and
chemical properties of the aromatic benzene ring can
therefore be investigated.[’-71 In the stable derivatives of
[6]paracyclophane 1, n = 6 , the benzene ring is bent by an
Fig. I. Molecular structure of 2
angle of a = 19.4-21.1 as determined by X-ray crystallography.14][5]Paracyclophanes exist only in solution at low
temperatures,[’] and, according to calculations, are bent by
an angle a of approximately 23 o.L61[4]Paracyclophanes are
reactive intermediates which can only be detected by trapO,
[*] Prof. Dr. M. W. Haenel, Dr. N. E. Blank [+], Prof. Dr. C. Kriiger,
Dr. Y:H. Tsay, H. Wientges
Max-Planck-Institut fur Kohlenforschung
D-4330 Miilheim a. d. Ruhr (FRG)
New address: Teroson GmbH, D-6900 Heidelberg (FRG)
0 VCH Verlagsgesellschafi mbH. 0-6940 Weinheim, 1988
To bend the “naphthalene bow” still further, we have
synthesized the [2](2,6)naphthalino[2]paracyclophane-1,11diene 3. Two Stevens rearrangements, accomplished by
reaction of 4 with dehydrobenzene (prepared from 1-(2’carboxyphenyl)-3,3-dimethyltriazene in refluxing xylene[”I) led to the ring-contracted
bis(pheny1thio)[2.2]phane 6 (isomeric
m.p. 128-135”C,
15-20%), which was oxidized to the disulfoxide 7 (>95%)
with m-chloroperbenzoic acid. Thermal elimination of
phenylsulfenic acid by gas-phase pyrolysis of 7 at 320°C
or by heating a solution of 7 in toluene to 160°C in a
sealed tube gave 3 together with diphenyl disulfide 8,
which is apparently formed from phenylsulfenic a ~ i d . l ’ ~ ~
Compound 8, which could not be separated from 3 by
chromatography, was reduced with excess threo-1,Cdimerin ethanol. Thiophecapto-2,3-butanediol (racemate) 911s1
no1 10, thereby formed, was removed by extraction with
aqueous sodium carbonate solution, and 4,5-dihydroxy-
0570-0833/88/0808-1064 $ 02.50/0
Angew. Chem. In[. Ed. Engl. 27 (1988) Nu. 8
Без категории
Размер файла
344 Кб
catalytic, methylated, isomerization, cyclooctadiene
Пожаловаться на содержимое документа