close

Вход

Забыли?

вход по аккаунту

?

N-Confused Porphyrins and Singlet Carbenes Is There a Connection.

код для вставкиСкачать
COMMUNICATIONS
matography on silica gel. Note: When more than 10 to 2 0 % of the amino ester
remained after the workup. the crude material was subjected to the reaction conditions again. After the second cycle the reaction was complete without a decrease in
the yield.
Received: November 9, 1994 127461 I€]
German version: A n g w . Chem. 1995. 107, 1095
Keywords: aza-Claisen reaction . azoninones . chirality transfer
. lactams
[I] a) R. Malherbe, D. BelluS, H e h . Chim. Actu 1978, 61. 3096; b) R. Malherbe,
G. Rist, D. BelluS, J Org. Ciiem. 1983, 48. 860.
[2] a) R. Ohrlein. R. Jeschke. B. Ernst, D. Bellus4, Erruhedron Lerr. 1989, 30.
3517; b) U. Nubbemeyer. R. Ohrlein, J. Gonda. B. Ernst. D. BelluS, Angew.
Chem. 1991. 103. 1533: Angeir.. Clrein. Inr. Ed. Engi. 1991, 30. 1465.
[3] a) G . Rossini. G . Spineti, E. Foresti. G. Pradella. J. Org. Chrtii. 1981,46, 2228:
) Ishida. H. Muramaru.
b) E. Vedejs, R. A. Buchanan. ihrd. 1984.49, 1 8 4 0 ; ~M.
S. Kato, Sj,nthesis 1989, 562.
[4] a) E. D. Edstrom, J. A m Chem. Soc. 1991,113,6690;b) M. M. Cid, E. PomboVillar, H c l v . Chint. ACIU 1993. 76. 1591: c) M. M. Cid. U . Eggnauer, H. P.
Weber. E. Pombo-Villar, Etrahedron Lrtt. 1991. 32, 7233.
[5] E. Edstrom [4al described the synthesis of a n hexahydroazoninone without a
chiral center.
[6] R. N. Icke, B. B. Wisegarver, G . A. Alles. Org. Synth. Coil. Vol. 3. 1955, 723.
[7] L. F. Tietze. T. Eicher. Rcwkrionen und Sjnrhesm nn Orgonise/i-Chrmrsc./ri,n
Pruktikiim. 2nd ed., Thieme. Stuttgart 1991, p. 135.
(81 E. J. Corey. A. Venkatesvarlu, J. A m . Clirm. SOC.1972. 94, 6190.
[9] All attempts at selectively reducing the ester function to avoid racemization in
the rearrangement failed; the resulting allyl alcohol was always contaminated
with 20 to 50% of the saturated compound. Because of the difficult separation
of these products and the low yield of the desired material this sequence was
abandoned. Cf. T. Moriwake, S:I. Hamano. S. Saito, S. Torii. Chem. Lrtt.
1987. 20135.
[lo] H. A. Hagemann. Org. Reoct. 1953, 7. 158,
[ l l ] The reaction conditions described by Edstrom [4a] afforded exclusively the von
Braun type products and decomposed material.
.
1954, 2006.
[I21 E. C. Bourne. M. Stacey, J. C. Tatlow. R. Worall, J. C h ~ iSol,.
1990.
[13] a) E. Valenti, M. A. Pericas, F. Serratosa. D. Maha, J. Chem. Res.
118; b) W. T. Brady, Y.-F. Giang. A. P. Marchand. A:H. Wu, J. Org. Chem.
1987. S2. 3457; c) R. J Clemens. J. A. Hyatt, ;hid 1985, SO, 2431: d) W. T.
Brady, R. A. Owens, Tetruhedron Lett. 1976, 1553; e) G . H. Olah, A.-H. Wu.
0. Farooq. SynIhtw.s 1989, 568.
[14] Only with dichloroketene could an analogous azoninone be isolated in low
yield. (BelluS et al. [1,2] obtained the same results wlth allyl sulfides.) a) W. T.
Brady, W. L. Vaughn, H . G. Liddell, J. Org. Chcm. 1966,3/. 626. b) M. Diederich. Diplomarbeit. Berlin 1992
[I51 E. Vedejs. M. Gringas. J. A m . Chcm. Soc. 1994, 116, 579.
[16] K. Soai, H. Oyamada, A. Okawa, Sjnth. Commun. 1982, 12. 463.
[I71 D. A. Dale. D. L. Dull. H. S. Mosher. J. Org. CIiem. 1969, 34. 2543.
[IX] a) F. E. Ziegler, Climi. REV.1988.88, 1429: b) W. S. Johnson, L. Werthemann,
W. R. Bartlett. T. J. Brocksom. T. Li, D . J. Faulkner, M. R . Petersen, J. Am.
Chem. S o ( . 1970, Y2, 741: c) D. Felix. K. Gschwend-Steen, A. E. Wick. A.
Eschenmoser, Helv. Chrm. A ~ t u1969, 52. 1030; d) T. Tsunoda, M. Sakai, 0 .
Sakai. Y. Sako, Y. Honda, S. Ito, Ezlrahedron L E U . 1992, 33. 1651.
[I91 a) R E. Ireland, R. H. Mueller, A. K . Willard, J. Am. Chew?. So[.. 1976, 98.
28613; b) R. E. Ireland, R. H. Mueller, ibid. 1972. Y4. 5897.
[20] 2: [4i2= -70.8 (c = 3.3, CHCI,): 3: [.It2 = -70.6 (c = 2.9. CHCI,); 5:
[a];’ = - 33.7 (C = 3.52. CHCI,): 6: [a]:’ = 38.9 (c = 8.1, CHCI,); 8 :
= 46.7 (C = 5.6. CHCI,); 9 : [XI;’ = 51.2 ( C = 2.0, CHCI,); 14:
[a]P = 86.3 (C = 3.1, CHCI,); 15: [XI;’ =17.0 (c =1.13. CHCI,); 16:
[a]:’ = - 116.1 (c = 3.6. CHCI,); 17a: [XI;’ = 15.0 (c = 1.6, CHCI,. Mosher
ester).
[21] B. Liining. C. Lundin, k t u Chern. Simid. 1967. 2 i . 2136.
-
-
N-Confused Porphyrins and Singlet Carbenes :
Is There a Connection?**
Abhik Ghosh*
The isolation of 2-aza-21 carba-5,10,15,2O-tetraarylporphyrins
as by-products of a standard preparative method for tetraarylporphyrins was an important and unexpected development in
porphyrin chemistry.“, 21 Interest in these novel compounds
(e.g. l ) , also dubbed N-confused porphyrins, centers on their
remarkable ability to act as tetracoordinate ligands and form
complexes (e.g. 2) with metal-carbon bonds (Scheme l).”]
. .
1
Scheme 1. Ar
= p-tolyl; a )
2
NiCI,.6H,O, CHClJEtOH, reflux, 0.5 h
However, neither the original report“] on this discovery nor a
critical commentaryr3]on these compounds proposed a rationale for the unexpectedly labile central C-H bond of 1.
Here we point out a striking analogy between the remarkable
carbon acidity of N-confused porphyrins and the properties of
completely different organic compounds, namely
mu
tB u
stable “bottleable” carbeI
I
nes such as 4 synthesized
by Arduengo et al. by the
deprotonation of N,N-diS
H
I
N
>
alkylimidazolium ions 3.[41
Thus, complex 2 may be
tB u
tBu
\
“ i
formally regarded as a
3
4
nickel@)-stabilized
carbene. In addition, we propose that the stability and aromaticity of the ligand in 2 may
account for the hitherto unexplained carbon acidity of 1 and its
N-protonated form, 1-H’.
Our proposal, if essentially correct, should have far-reaching
implications for the electronic structure and chemical reactivity
of N-confused porphyrins. Accordingly, we carried out highquality ab initio[’] and local density functional (LDF)r61calculations on compounds 5 - 9 to test the validity of the proposed
analogy as well as to develop the first theoretical picture of the
electronic structure of N-confused porphyrins. Compounds 5
and 8 are the unsubstituted analogues of 1 and 2, respectively.
Species 6 is the free base form of the ligand in complex 8 and the
carbenic tautomer of 5. The zinc complex 7 was selected for
studying the interaction of the carbenic ligand 6 with a closed[*] Dr. Ghosh
Department of Chemistry and Minnesota Supercomputer Institute (MSI)
University of Minnesota
Minneapolis M N 55455 (Chemistry) and 55415 (MSI) (USA)
Present address:
Department of Chemistry, University of California
Riverside, CA 92521 (USA)
Telefax: Int. code (909)787-4713
e-mail: mf10105(~1~sc.msc.edu
or mf10105(a sk.msc.edu.
Thls work supported by the MSI and Prof. Jan Almlof.
+
[**I
1028
0 VCH
Verlugsgesellschafi m b H . 0-69451 Weinhebn, 1995
0570-0833iY5l0909-io2x $ lO.OU+ ,2510
Angeir. t h e m . In/.Ed. Enni. 1995. 34. No. 9
COMMUNICATIONS
5
6
7
0
and 9 corresponds to o ionization, whereas the fourth lowest
HF-KT IP of 6 corresponds to a o-ionized final state. L D F
calculations (using the ASCF method) also show that the lowest
o-IP of 6 is relatively low at approximately 7.31 eV compared to
the lowest x-IP of 6.65 eV and to the lowest o-IPs of 5 and 9 (for
which the HF-KT values listed in Table 1 may serve as estimates). An L D F calculation on the lowest energy o-ionized 'A'
state of 6 shows that the majority ( E 80 YO)of the unpaired spin
density is localized in the region of the carbenic lone pair. This
is indeed what one expects of a stable nucleophilic carbene of the
type synthesized by Arduengo et a1.[I2]For instance, the lowest
0-IP of 4, corresponding largely to ionization from the carbenic
lone pair, is 7.68 eV.["' which is quite close to the L D F value of
7.31 eV for 6. One difference between 4 and 6 is that the aromatic framework of 6 may be regarded as a large annulene whose
highest occupied x MOs have rather higher orbital energies (i.e.
lower HF-KTs) than the x HOMOS of imidazole. Thus, it is
understandable that the lowest L D F IP of 6 (7.31 eV. which
corresponds to a TI ionization) is significantly smaller than the
lowest IP (7.68 eV) of 4,which corresponds to o ionization.
Conceptually, the lowest A E S - Tvalues of 5-8 are not directly
comparable to the AES-T values of known carbenes, since the
lowest triplet states of 5-8 do not arise through electronic excitation from the carbenic lone pair, but rather have have xlx*'
configurations. The L D F values of the lowest A E , _ , values of
6-8 range from 26.4 to 31.0 kcalmol-'. As Fable 2 shows, the
9
shell Lewis acid such as a Z n 2 + ion. L D F calculations on 5-8,
including geometry optimizations, used double-zeta plus polarization (DZP) basis sets, the von Barth-Hedin exchangecorrelation functional,['] and a C, symmetry constraint. Ab initio calculations using Hartree-Fock ( H F ) and second-order
Merller - Plesset perturbation theories were carried out on LDFoptimized geometries of the organic molecules 5 and 6 and the
unsubstituted free base porphyrin 9 with generally contracted
double-zeta (DZ) basis setsts, 91 and direct electronic structure
methods["] that obviate the need for storing prohibitive volumes of electron repulsion integrals. In the optimized geometry
of 6 the C-C-C angle subtended at the carbenic center is only
104.2", which is extremely smal1 for dicoordinate carbon and
highly characteristic of singlet carbenes. Table 1 presents the
calculated Hartree-Fock orbital energies of the six highest occupied molecular orbitals (MOs) of the purely organic compounds 5. 6, and 9. Table 2 presents calculated L D F values of
the lowest singlet-triplet energy gaps ( L I I Z - ~for
) both the lowest x'x*' and olx*' triplet states, and the lowest o and x ionization potentials (IPS) of 5-8.
Tdhk 1 . Sign-reversed HF orbital energies [eV] of the six highest occupied MOs of
5. 6. and 9.
5
6
9
10.41 ( n )
10.31 ( n )
9.55 (n)
8.99 ( 8 )
6.66 (n)
6.40 (x)
9.74 (x)
9.11 (n)
8.54 (0)
8.44 (n)
6.54 (n)
5.63 (x)
10.2s (n)
10.20 (x)
9.30 (x)
9.07 (x)
6.52 (x)
6.20 (x)
Table 1 shows that the values of the H F orbital energies of 5
and 9 are extremely similar,[*'] at least for the six lowest IPS
(according to Koopmans' theorem, KT), while the values of 6
are considerably different. Thus, none of the six lowest IPS of 5
Table 2. LDF values of AEs
Molecule
[kcalmol-'1 and lowest a-and PIPS [ev].
I.owe?t [P
Lowest AEs
a'x*' triplet
a
x
7.22
6.65
6.76
7.72
dx*' triplet
5
[a1
[a1
la1
6
I
26 4
29.6
31.0
41.2
34.9
56 3
7.31
6.96
7.79
8
[a] These values are not available since the calculations o n thc pertinent. highly
excited states could not he converged.
lowest d x * ' triplet states of 6-8 are even higher in energy.
These AES-Tvalues may be compared to calculated values of
45 kcalmol-' for CF,,[l3I a prototypical singlet carbene, and
79.4 kcal mol- ' for unsubstituted imidaz01-2-yIidene.['~]Thus,
the AES-= values of 6-8, while substantial, are significantly
lower than those of the stable imidazol-2-ylidene-type CdrbeneS.
Once again, this is a consequence of the relatively low energy of
the lowest ~ l l n *states
~ of large-ring aromatic compounds such
as porphyrins.
Although 6 exhibits the salient structural and electronic attributes of a stabilized singlet carbene, it is still higher in energy
than 5 by a substantial margin-30.4, 32.8, and 42.2 kcalmol-'
at the LDF, HF, and MP2 levels, respectively.[' 'I In view of the
significant energy difference between tautomers 5 and 6, how
can we account for the activation of the central C - H bond in 1
(the known analogue of 5) by a reagent as mild as NiCI,? To
probe this issue, we examined the interaction of 6 with a Z n Z +
ion (a closed-shell Lewis acid) and with a Ni" ion by performing calculations on 7 and 8, respectively.['61 Table 2 shows that
the lowest o-IPS of 6 and 7 are similar, as are the energies of the
lowest dx*' triplet states relative to the ground states. However, unlike the lowest cationic 'A' (i.e. a-ionized) state of 6, the
unpaired electron spin in the lowest 'A' state of 7 is delocalized
rather evenly on the carbenic carbon, the three central nitrogen
atoms, and the zinc ion, suggesting a certain stabilization of the
carbenic lone pair by covalent interactions with the zinc center.
COMMUNICATIONS
The results on 8 are drastically different from those on 6 and 7.
Thus, the energies of both the lowest dn*' triplet and cationic
'A' states of 8, relative to the neutral ground state, are much
higher than the corresponding values for 6 and 7. Indeed, in
the lowest energy cationic 2A' state of 8, the nickel ion carries
nearly all the unpaired spin density, leaving negligible spin
density at the carbenic center. In other words, the carbenic
lone pair does not contribute to the highest occupied o MO
of 8, but to an MO with an even lower orbital energy. An
examination of the high-lying CJ MOs of 8 reveals that the
stability of 2 or 8 stems significantly from a highly favorable
covalent interaction between the carbenic lone pair and the
empty d, atomic orbital of an NiZ+ion (d') placed in a strong
square-planar ligand field as well as from a metal-ligand n
bond.
Both Arduengo-type carbenes and the carbenic tautomers of
N-confused porphyrins feature a carbenic lone pair in the nodal
plane of a highly stable aromatic n: system. This stereoelectronic
feature results in a substantial singlet-triplet energy gap. As in
the case of Arduengo-type carbenes, significant fractions of the
unpaired spin density in the lowest o-ionized states of 6 and 7
reside on the carbenic center. The lowest CJ IPS of 6 and 7 are
also close to the lowest o IPS of the imidazol-2-ylidene-based
carbenes. Thus, our calculations lend strong support to the proposal that 6 is indeed an Arduengo-type aromatic carbene. The
specially favorable covalent interactions between 6 and the
Ni2+ ion
'
provides additional insight into the surprising ease
with which N-confused porphyrins form metal-carbon
bonds.
Received: November 24. 1994 [Z7496IE]
German version: Angew. Chem. 1995. 107, 11 17
Keywords: ab initio calculations carbenes . LDF calculations .
porphyrinoids
~
[l] P. J. Chmielewski, L. Ldtos-Craiynski. K. Rachlewicz, T. Glowiak, Angew.
Chem. 1994. 106, 805; Angew. Chem. lnt. Ed. Engl. 1994, 33. 779.
[2] H. Furuta. T. Asano, T. Ogawa, J. Am. Chem. Sor. 1994, 116, 767.
[3] J. Sessler. Angen. Chem. 1994,106,1410; Angen. Chem. Int. Ed. Engl. 1994,33,
1348.
[4] a) A. J. Arduengo, R. L. Harlow, M. Kline, J. Am. Chem. Soc. 1991, 113,361;
b) A. J. Arduengo 111, H.V. R. Dias, R. L. Harlow. M. Kline, ibid. 1992, f14,
5530; c) N. Kuhn, T. Kratz, Synthesis 1993, 561.
[5] Computer program used: J. Almlof, K. Faegri, M. W. F. Feyereisen. T. Fischer,
K. Korsell, H. P. Liithi: DISCO, a direct SCF and MP2 code
[6] The LDF calculations were carried out with the program DMOL. using methods described in a) B. Delley, J. Chrm. Pliys. 1990, 92. 508; b) ibid. 1991, 94,
7245.
[7] U. von Barth, L. Hedin, J. Phy.?. C\Solid Stale Phys. 1972, 5 , 1629.
[8] General contraction: a) R. C. Raffeneti. J. Chem. Phjs. 1973, 58, 4452:
b) M. W. Schmidt; K. Ruedenberg. rbrd. 1979, 71, 3951.
(91 The basis sets for C , N. and H were (6s3p)/[3s2pj, (6s3p)/[3s2p13 (3s)/[2s],
respectively, and were obtained from F, B. vanDuijneveldt, IBM Research
Report RJ945, 1971. The hydrogen exponents were multiplied by a scaling
factor of 1.44
(101 a) Direct SCF calculation: J. Almlof, K. Faegri, K. Korsell, J. Comput. Chem.
1982,3, 385; b) Direct MP2 calculation: S. Saebc?,J. Almlof. Chem. Phys. Lett.
1989, 154, 521.
[ l l ] For recent a b initio calculations on porphyrinic molecules: a) A. Ghosh, J.
Almlof, P. G . Gassman, Chem. Phys. Lett. 1991, 186, 113; b) P. G. Gassman,
A. Ghosh. J. Almlof, J. Am. Chem. Soc. 1992,114,9990:c) A. Ghosh. J. Almlof,
Chrm. Phyx. Lett. 1993, 213, 519; d) J. Almlof, T. Fischer, P. G . Gassman, A.
Ghosh, M. HSser, J. Phys. Chem. 1993, 97, 10964; e) A. Ghosh, P. G. Gassman, J. Almlof. J. Am. Chem. Sac. 1994, 116. 1932: f ) M. Merchin, E. Orti, B.
ROOS,Chem. Phys. Lett. 1994,221, 136; g) ibid. 1994, 226,27; h) A. Ghosh, J.
Phys. Chem. 1994. 98. 11 004; i) A. Ghosh, J. Fitzgerald. P. G. Gassman, J.
Almlof, Inorg. Chem. 1994. 33, 6057; j ) A. Ghosh. J. Almlof, J. Phy.?. Clzrm.
1995, 99. 1073. A. Ghosh, J. Am. Chem. Sur. 1995, 117, in press.
1030
8 VCH
Verlugsgssellschaft mbH, 0-69451 Weinheim. 1995
[12] A. J. Arduengo 111, H. Bock, H. Chen. M. Denk, D. A. Dixon, J. C. Green,
W. A. Herrmann, N. L. Jones, M. Wagner, R. West, J Am. Chem. Soc. 1994,
116, 6641.
[13] H. F. King, A. Komornicki, J Chem. Phys. 1986, H4, 5465.
(141 D. A. Dixon, A. J. Arduengo 111, J. Pkjs. Chem. 1991, 95, 4180.
[IS] At the LDF level compound 5is roughly 16.5 kcalmol-' higher in energy than
9. This energy difference should be regarded as an upper limit, since the optimization of 5 used a C, symmetry constraint, which is an approximation.
[16] For an extensive list of reference on Lewis acid complexes of Arduengo-type
carbenes, see: A. J. Arduengo 111, H. V. R. Dias. D. A. Dixon. R. L. Harlow,
W. T. Klooster, T. F. Koetzle. J. Am. Chem. Sor. 1994, 116, 6812.
Silicon Analogues of Grignard Compounds :
Synthesis and Structure of Amine-Stabilized
Trimethylsilylmagnesium Halides**
Richard Goddard, Carl Kruger, Nazmi A. Ramadan,
and Alfred Ritter*
Dedicated to Professor Hubert Schmidbauv
on the occasion of his 60th birthday
For a long time it has been suspected*'] that triorganosilyl
Grignard compounds play a role as intermediates in the reaction
of haIo(triorgano)silanes with elemental magnesium to give
hexaorganodisilanes. However, all attempts to synthesize members of this class of compounds in analogy to the classical methods of preparing carbon Grignard reagents have been unsuccessful.
Indirect evidence for the existence of triorganosilylmagnesium
halides in ether has been recently obtained by Oehme and coworkersr2]during the metal exchange reaction of tris(trimethylsi1yl)silyllithium (Me,Si),SiLi. 3 T H F with magnesium bromide. The assumed intermediate, tris(trimethylsily1)silylmagnesium bromide (Me,Si),SiMgBr, was not isolated but immediately
underwent further reaction in ether solution.
We have now synthesized Grignard analogous trimethylsilylmagnesium compounds easily as complexes with amine chelating ligands, directly from halo(trimethy1)silanes and magnesium, using the highly reactive pyrophoric magnesium (Mg*) of
BogdanoviC et al.[,]
So far we have observed that only bromo(trimethy1)silane and
iodo(trimethy1)silane and not the usual chloro(trimethy1)silane
react under our chosen reaction conditions to give trimethylsilyl
Grignard compounds. In contrast to the classical Grignard synthesis, the reaction cannot be carried out in oxygen-containing
solvents such as tetrahydrofuran or diethyl ether because of
quantitative ether cleavage during the course of the reaction.
Hence the reactions were carried out in dried toluene in the
presence of tetramethylethylenediamine or pentamethyldiethylenetriamine as complexing ligands.
~
[*I
Dr. A. Ritter, Prof. Dr. N. A. Ramadan[+]
Max-Planck-Institut fur Strahlenchemie
Postfach 10 13 65, D-45413 Mulheim an der liuhr (Germany)
Telefax: Int. code + (208)306-3951
Dr. R. Goddard, Prof. Dr. C. Kruger
Max-Planck-Institut fur Kohlenforschung
D-45470 Miilheim an der Ruhr (Germany)
[ '1 Permanent address: Egyptian Petroleum Research Institute
Nasr City, 7th Region. Cairo (Egypt)
[**I We thank Prof. Dr. M. Demuth and Dr. J. Leitich for valuable discussions, Mr.
P. Bayer and Mr. K. H. Claw for experimental assistance, as well as Prof. Dr.
K. Schdffner for his kind support. N. A. R. thanks the DFG for financial
support for his stay at the MPI fur Strdhlenchemie in 1992 and 1994.
0570-OH33195j0909-1030$10.00
+ .25/0
Angrn. Chem. Int. Ed. Engl. 1995, 34, No. 9
Документ
Категория
Без категории
Просмотров
0
Размер файла
374 Кб
Теги
connection, carbene, single, porphyrio, confused
1/--страниц
Пожаловаться на содержимое документа