Synthesis of N-confused porphyrin analogues by β-azafulvenone tetramerization

Article abstract of DOI:10.1021/jo9609438

β-Azafulvenones 16 and 33 were generated from 14 or 18 and 31 or 37, respectively, and directly observed by IR spectroscopy. These heterocyclic ketenes tetramerize to the macrocycles 20 and 34.

Full text of DOI:10.1021/jo9609438

J . Org. Chem. 1996, 61, 8125-8131
8125
Syn th esis of N-Con fu sed P or p h yr in An a logu es by â-Aza fu lven on e
Tetr a m er iza tion
Greg GuangHua Qiao, Ming Wah Wong, and Curt Wentrup*
Department of Chemistry, The University of Queensland, Brisbane, Queensland 4072, Australia
Received May 21, 1996X
â-Azafulvenones 16 and 33 were generated from 14 or 18 and 31 or 37, respectively, and directly
observed by IR spectroscopy. These heterocyclic ketenes tetramerize to the macrocycles 20 and
34.
In tr od u ction
work13 has shown that in heterocyclic ketenes possessing
an R-nitrogen atom (R-azafulvenones) such as 2-carbonyl-
2H-indole (6) and 2-carbonyl-2H-pyrrole (8), this nitrogen
atom acts as a nucleophile toward the ketene moiety of
another molecule, thus leading to the dimers diindolo-
[1,2-a;1′2′-d]pyrazine-8,16-dione (7) and dipyrrolo[1,2-
a;1′2′-d]pyrazine-5,10-dione (9), respectively (Scheme 1).
It was the purpose of the present study to investigate
the behavior of heterocyclic ketenes 10 (X ) hydrogen
or carbocyclic or heterocyclic ring) possessing a â-nitrogen
atom (â-azafulvenones). Instead of dimerization, oligo-
merization should lead to the macrocycles 11. On the
basis of ring strain arguments, the tetramer may be
expected to be the most stable compound.14 Such tet-
ramers (20 and 34) will have four C-H bonds inside the
macrocyclic ring and thus constitute “tetra-N-confused”
porphyrin analogues.
Porphyrins are important macrocycles for most organ-
isms.1 Lately, there has been considerable interest2,3 in
the preparation of non-natural analogues and isomers of
the porphyrin system, including porphycene (2),4 corr-
phycene (3),5 hemiporphycene (4),6 benziporphyrins,7 and
“N-confused”8 porphyrins 5. The isomers 5, synthesized
by the groups of Furuta8 and Latos-Grazynski,9 are
particularly interesting, as one of the pyrrole subunits
has been rotated (“confused” or “inverted”) with respect
to its normal position in conventional porphyrins. The
inner C-H bond in 5a is readily metalated with the
formation of a Ni(II) metal complex involving three
pyrrole nitrogens and C-3 of the fourth ring.9b,10,11
Resu lts a n d Discu ssion
1. Gen er a tion a n d Obser va tion of 3-Ca r bon yl-
3H-in d ole. (a ) Fr om 3-Dia zo-4-oxo-3,4-d ih yd r oqu in -
olin e (14). 3-Nitro-4-hydroxyquinoline was reduced
to 3-amino-4-hydroxyquinoline (12) according to Su¨s’
method.15 Diazotization of 12 afforded 4-oxo-3-diazo-3H-
dihydroquinoline (14)15 after neutralization of the dia-
zonium salt 13. The diazo compound 14, purified by
vacuum sublimation, was then sublimed onto a 12 K
window with Ar matrix isolation and subsequently pho-
tolyzed at 254 nm. The photolysis caused elimination of
nitrogen, formally to give an R-oxo carbene 15, which
underwent Wolff rearrangement to 3-carbonyl-3H-indole
(16) (Scheme 2). Ketene 16 has a very strong peak at
2140 cm-1. In the fingerprint region, the main peaks are
at 1473, 1448, and 1459 cm-1. The reacted diazo com-
pound 14 features a diazo absorption at 2110 cm-1 and
It is well known that simple ketenes prefer to dimerize
by way of a [2+2] cycloaddition reaction.12 Our previous
a carbonyl absorption at 1662 cm-1 14a
The experimental
.
X Abstract published in Advance ACS Abstracts, October 15, 1996.
(1) (a) Biochemistry: The Porphyrins; Dolphin, D., Ed.; Academic
Press: New York, 1978; Vol. VI, Part A; Vol. VII, Part B. (b) Heme
and Hemoproteins; de Matteis, F., Aldrige, W. N., Eds.; Springer:
Berlin, 1978.
IR spectrum of 16 is in very good agreement with the
BLYP/6-31G* calculated spectrum (Figure 3 in the sup-
porting information).
(2) (a) Sessler, J . L. Angew. Chem., Int. Ed. Engl. 1994, 33, 1348.
(b) Borman, S. Chem. Eng. News 1995, J une 26, 30. (c) For “artificial
porphyrins” that are not conjugated 18-π systems but different oxidized
forms of porphyrinogens, see: Floriani, C. Chem. Commun. 1996, 1257.
(3) Vogel, E. Pure. Appl. Chem. 1990, 62, 557. Vogel, E. Ibid. 1993,
65, 143.
(4) Vogel, E.; Ko¨cher, M.; Schmickler, H.; Lex. J . Angew. Chem., Int.
Ed. Engl. 1986, 98, 262; Angew. Chem., Int. Ed. Engl. 1986, 25, 257.
(5) Sessler, J . L.; Brucker, E. A.; Weghorn, S. J .; Kisters, M.; Scha¨fer,
M.; Lex, J .; Vogel. E. Angew. Chem., Int. Ed. Engl. 1994, 33, 2308.
(6) Callot, H. J .; Rohrer, A.; Tschamber, T. New J . Chem. 1995, 19,
155.
(10) For other efforts to synthesize C-coordinated porphyrin ana-
logues, see: Kernbach, U.; Ramm, M.; Luger, P.; Fehlhammer, W. P.
Angew. Chem., Int. Ed. Engl. 1996, 35, 310.
(11) For theoretical calculations relating to the complexation of 5
by Ni2+, see: Ghosh, A. Angew. Chem., Int. Ed. Engl. 1995, 34, 1028.
(12) (a) Hyatt, J .; Raynolds, P. W. Org. React. 1994, 45, 159. (b)
Wentrup, C.; Heilmayer, W.; Kollenz, G. Synthesis 1994, 12, 1219. (c)
Tidwell, T. T. Ketenes; Wiley-Interscience: New York, 1995.
(13) (a) Qiao, G. G.; Meutermans, W.; Wong, M. W.; Tra¨ubel, M.;
Wentrup, C. J . Am. Chem. Soc. 1996, 118, 3852. (b) Gross, G.; Wentrup,
C. J . Chem. Soc., Chem. Commun. 1982, 360.
(7) Lash, T. D. Angew. Chem., Int. Ed. Engl. 1995, 34, 2533.
(8) Furuta, H.; Asano, T.; Ogawa, T. J . Am. Chem. Soc. 1994, 116,
767.
(9) (a) Chmielewski, P. J .; Latos-Grazynski, L.; Rachlewicz, K.;
Glowiak, T. Angew. Chem., Int. Ed. Engl. 1994, 33, 779. (b) Chmielews-
ki, P. J .; Latos-Grazynski, L. J . Chem. Soc., Perkin Trans. 2 1995, 503.
(14) (a) Preliminary communication on 16 and 20: Qiao, G. G.;
Wentrup, C. Tetrahedron Lett. 1995, 22, 3913. (b) For the structure
and a different synthesis of 20, see: Oliver, J . E.; Lusby, W. R. J .
Heterocycl. Chem. 1991, 28, 1565.
(15) Su¨s, O.; Glos, M.; Mo¨ller, K.; Eberhardt, H. D. J ustus Liebigs
Ann. Chem. 1953, 583, 150.
S0022-3263(96)00943-7 CCC: $12.00 © 1996 American Chemical Society

8126 J . Org. Chem., Vol. 61, No. 23, 1996
Qiao et al.
Sch em e 1
Sch em e 3
18 can also be matrix isolated in the presence of CO. In
this case, photolysis at 254 nm gave rise to the same
ketene 16 as observed by IR spectroscopy.14a Although
the ketene region around 2100 cm-1 was obscured by the
large CO absorption, the fingerprint absorptions at 1473,
1448, and 1062 cm-1 clearly demonstrate the formation
of ketene 16. The diazo compound 18 was consumed
during the photolysis. Obviously, photolysis caused the
elimination of nitrogen from 18, forming carbene 19,
which was then immediately trapped by CO in the matrix
to generate ketene 16 (Scheme 2).17
2. Tetr a m er iza tion of 16 to 1,3′-tetr a k is(in d olyl-
m eth a n on e) (20). Isolation of ketene 16 at 12 K and
subsequent warmup revealed that this ketene is only
stable below 58 K. This instability is due to the imine
function which acts as a nucleophile toward another
ketene molecule.18 IR spectroscopy showed that the
product of this thermal matrix reaction is a carbonyl
compound. The same product can be obtained prepara-
tively in 75% yield by FVT of diazo compound 14 at 550
°C. When the product of the 58 K reaction was warmed
to rt and dissolved in DMSO-d6, it was shown by 1H NMR
spectroscopy to be identical with the product from
preparative FVT. The molecular mass of 529 amu by
mass spectrometry indicates a tetramer of ketene 16
(Scheme 3). In the mass spectrum, ions due to one, two,
three, and four monomer units are seen. Both 1H and
13C NMR demonstrate that the tetramer is a symmetrical
molecule with four identical indolecarboxamide subunits.
This tetramer was therefore assigned as 1,3′-tetrakis-
(indolymethanone) (20).14 The NMR signal assignment
was confirmed by two-dimensional 13C-1H correlations
(COLOC) with both short- and long-range couplings.
MM3 calculations of the structure indicate that 20 is the
most stable oligomer compared to the trimer, pentamer
or hexamer.19 The tetramerization of 16 to 20 was
directly monitored at T e 70 K by IR spectroscopy (Figure
5 in the supporting information).
Sch em e 2
Tetramer 20 forms a light-yellow powder with melting
point above 290 °C. It is insoluble in most organic
solvents. FVT of 20 at 700 °C regenerates ketene 16 as
the only product as evidenced by the strong IR spectrum
of the Ar matrix isolated material (Scheme 3).
Flash vacuum thermolysis (FVT) of the same R-oxo
diazo ketone 14 at 650 °C gave the same ketene, isolated
in Ar matrix at 12 K.
(b) F r om 3-Dia zo-3H-in d ole (18). Treatment of
indole with amyl nitrite in absolute EtOH/NaOEt af-
forded 3-oximino-3H-indole (17).16 This oxime 17 was
dissolved in degassed NaOH solution together with
concentrated ammonia and treated with NaOCl. 3-Diazo-
3H-indole (18) thus formed is stable in ether solution at
room temperature for several days. It can be sublimed
onto a 12 K deposition window under vacuum with Ar
as a carrier gas. The Ar matrix isolated 18 has a very
strong diazo absorption at 2092 cm-1 together with a
strong absorption at 1429 cm-1 in the fingerprint region.
3. Oth er P er cu sor s for Tetr a m er 20. We have
reported13a that under FVP conditions methyl indole-
(17) For other examples of carbene trapping by CO in matrices,
see: Sander, W. Chem. Rev. 1993, 93, 1583.
(18) The reaction between ketenes and pyridine gives directly
observable ylides: Qiao, G. G.; Andraos, J .; Wentrup, C. J . Am. Chem.
Soc. 1996, 118, 5634.
(19) Force field calculations performed by Dr. D. J . Brecknell with
the MM3(92) molecular mechanics program (Allinger, N. L.; Yuh, Y.
H.; Lii, J .-h. J . Am. Chem. Soc. 1989, 111, 8551. Lii, J .-h.; Allinger, N.
L. J . Am. Chem. Soc. 1989, 111, 8566, 8576). Calculated strain energy
per monomer unit, for tetramer: Es ) 32.6 kcal/mol; trimer: Es ) 45.7
kcal/mol; pentamer: Es ) 35.7 kcal/mol; hexamer: Es ) 35.8 kcal/mol.
(16) Madelung, W. J ustus Liebigs Ann. Chem. 1914, 504, 58.

Synthesis of N-Confused Porphyrin Analogues
J . Org. Chem., Vol. 61, No. 23, 1996 8127
Sch em e 4
Sch em e 5
Ketene 16 then tetramerized to 20. It should be pointed
out that these are not FVT reactions since the tetramer
was isolated from the residue of the tube containing the
starting material. Furthermore, heating of 27 or 28 at
ca. 200 °C in a Kugelrohr distillation apparatus under
vaccum also gave tetramer 20, isolated from the solid
residues.
Direct evidence for the formation of ketene 16 from
28 was, however, obtained by IR spectroscopy of the
product from a 940 °C FVT of 28 with Ar matrix isola-
tion at 12 K. 16 was identified by comparison with
the spectrum shown in Figure 3 (in the supporting infor-
mation). However, at this high FVT temperature, a
strong absorption at 2127 cm-1 was also seen. The latter
is due to the formation of ketene 24, the product of
migration from position 3 to position 2 of the indole ring
followed by retro-Wolff rearrangement (Scheme 4).13a The
possibility that 24 is formed by direct ring opening of
ketene 16 was ruled out in FVT experiments with 14.
FVT of 14 at 400 to 1000 °C and subsequent Ar matrix
isolation of the products at 12 K afforded ketene 16 as
the only product according to IR spectroscopy. No
absorption due to ketene 24 (2127 cm-1) could be de-
tected.
3-carboxylate (21) rearranges to methyl indole-2-carboxy-
late (22), which subsequently eliminates methanol to
form 2-carbonyl-2H-indole (6). Ketene 6 is highly un-
stable and undergoes either dimerization to 7 or retro-
Wolff rearrangement to cyanoketene 24. Ketene 24 is
trapped by MeOH to give the ester 25 as the final product
(Scheme 4).13a In the matrix isolation work, we observed
a small absorption ascribable to ketene 16, formed by
direct elimination of MeOH from ester 21 before any ester
group migration takes place. Because the amount of
ketene 16 is so small, the tetramer 20 is not isolable from
a preparative FVT of 21.
It was thought that anhydrides such as 27 and 28
might be more useful precursors of ketene 16. Anhydride
27 (Scheme 4) can be easily obtained in quantitative yield
by stirring indole-3-carboxylic acid in acetic anhydride
solution. Preparative FVT of 27 at 500 °C gave 80% of
indole-3-carboxylic acid, together with 10% of a mixture
of indole and 3-acetylindole in the ratio of 3:2 (GC-MS).
On raising the temperature to 750 °C, the yield of indole
increased to 85%, and 10% of tetramer 20 was isolated
from the solid residue at the entrance to the oven. The
tetramerization takes place in the solid as shown in the
following experiment.
Compared to diazo compound 14, anhydride 28 is much
easier to synthesize and high yields of tetramer 20 can
be achieved from this simple precursor.
4. Gen er a tion a n d Obser va tion of 3-Ca r bon yl-
3H-p yr r ole (33). 4-Hydroxy-3-aminopyridine (29), ob-
tained by reduction of 3-nitro-4-hydroxypyridine with
Raney Ni in EtOH solution, was diazotized with iso-
amyl nitrite in dry EtOH containing HCl. The diazonium
salt 30 thus obtained was stirred with solid sodium
carbonate in dry ether to give 4-oxo-3-diazopyridine (31),
which is quite stable in ether solution (Scheme 5). Ether
was evaporated and the isolated solid 31 was quickly
placed in the cryostat deposition tube. Compound 31 is
not stable as a neat solid, and part of it decomposed
during this transfer operation. Compound 31 was then
sublimed in the cryostat system at 60 °C (10-6 mbar)
with Ar or N2 as carrier gas for deposition on a 12 K
BaF2 window. The matrix isolated 31 showed a typical
strong diazo absorption at 2130 cm-1 (Ar) or 2144 cm-1
(N2). 31 was subsequently subjected to 254 nm UV
photolysis for 5 min. Under these conditions, nitrogen
is eliminated and an oxocarbene 32 is probably formed.
A rapid Wolff rearrangement transforms 32 to 3-carbo-
nyl-3H-pyrrole (33). In Figure 1, the positive spectrum
(b) is due to the formed ketene 33, and the negative
spectrum the reacted diazo compound 31. Ketene 33 has
a very strong CdC)O stretching absorption at 2146 cm-1
(Ar matrix; 2150 cm-1 in N2 matrix) together with
Better results were obtained with the anhydride of
indole-3-carboxylic acid (28), prepared by reaction of
indole-3-carboxylic acid with DCC/Et3N in acetone solu-
tion.20 FVP of 28 at 700 °C afforded 87% of tetramer 20
with indole as a byproduct, thus indicating that ketene
16 and indole-3-carboxylic acid were formed initially.
(20) Suvorov, N. N.; Golubev, V. E. Pharm. Chem. J . 1967, 1, 440.

8128 J . Org. Chem., Vol. 61, No. 23, 1996
Qiao et al.
In analogy with the indole case (vide supra), it was
thought that the anhydrides 35 or 36 could be useful
precursors. Both 35 and 36 were synthesized and
subjected to preparative FVT investigation. Unlike the
indole case, unfortunately, FVT of 35 and 36 gave none
of the tetramer 34. In search of a better thermal
precursor of 33, methyl 4-hydroxynicotinate (37) was
chosen and synthesized according to known procedures.21
F VT of 37 in th e Low -Tem p er a tu r e Ra n ge. Isola -
tion of 39. Thermal reaction of ester 37 in the temper-
ature range 500-700 °C is expected to cause elimination
of methanol, forming R-oxo ketene 38. It is also know
that oxo ketenes usually dimerize in a specific [2 + 4]
cycloaddition manner.12b When ester 37 was subjectd to
FVT at 560 °C, a new compound, 10-oxo-10H-pyrano[3,2-
c:5,6-c′]dipyridine (39) was isolated in 61% yield. The
specific formation of 39 is readily explained in terms of
[2 + 4] cyclodimerization as shown in Scheme 7. The
initial dimer 40 undergoes ring opening to an imino-
ketene 41. The latter recyclizes to 42, which under FVT
conditions eliminates CO2 to afford the stable product 39.
The ring opening of another spirodienone derivative to a
ketene, which subsequently cyclizes and eliminates CO2
to give a dibenzofuran derivative has been reported by
Sugawara et al.22
F igu r e 1. (a) Calculated IR spectrum of ketene 33 (BLYP/6-
31G*; unscaled). (b) Positive spectrum: formed spectrum from
photolysis of 31 with 254 nm lamp for 2 min; negative
spectrum: reacted diazo compound 31 isolated in N2 matrix
at 12 K.
Sch em e 6
Compound 39 is a symmetrical molecule. In the 1H
NMR spectrum, the signal due to H(C-1) and H(C-9)
appear at a very low field (9.35 ppm). This is typical for
xanthone-type heterocycles.23,24 It is interesting to note
that the unsymmetrical compound 46,23 which would
have resulted from the alternative dimer 43 in a similar
manner, was not detectable by NMR of the FVT product.
This is in agreement with orientations usually observed
in R-oxo ketene dimerizations.12b
F VT of 37 in th e High -Tem p er a tu r e Ra n ge. Iso-
la tion of 34. When ester 37 was subjected to FVT at
1000 °C, tetramer 34 was isolated in 90% yield without
any 39 being produced. In this experiment, a liquid N2
cooled U-tube was used as a cold trap. This causes the
methanol and ketene to condense in different parts of
the trap, thus avoiding their recombination. The result
clearly indicates that, under these severe thermolysis
conditions, the first-formed R-oxo ketene 38 underwent
further elimination of CO and Wolff rearrangement to
ketene 33, the latter cyclizing to tetramer 34 in good
yields in the absence of other nucleophiles (Scheme 6).
This thermal reaction is very clean, and tetramer 34 is
the only product isolated in high yield (90%). On FVT
of 37 at 1100 °C with Ar matrix isolation of the product,
ketene 33 together with peaks due to methanol and a
large absorption due to CO were identified by IR spec-
troscopy.
another three weak peaks in the fingerprint region. The
observed IR spectrum is in very good agreement with the
ab initio calculated spectrum (Figure 1a) at the BLYP/
6-31G* level. Electron correlation at the MP2 level gave
similar results, but with better intensity agreements. The
MP2/6-31G* spectrum is shown in the supporting infor-
mation.
5. Tetr a m er iza tion of Keten e 33 to 1,3′-Tetr a k is-
(p yr r olylm eth a n on e) (34). Preparative FVT of diazo
compound 31 at 550 °C yielded a light brown compound
with a molecular weight of 372 amu. In analogy with
the indole case, this compound is considered to be a
tetramer of ketene 33, namely 1,3′-tetrakis(pyrrolyl-
methanone) (34) (Scheme 6). Because of the instability
of 31, the yields of preparative FVT reactions were,
however, extremely low (∼1%). Hence, an alternative
precursor was needed for further investigation.
(21) Ross, W. C. J . J . Chem. Soc. C 1966, 1816. Kitagawa, T.;
Satoshi, M.; Hirai, E. Chem. Pharm. Bull. 1978, 26, 1403.
(22) Izuoka, A.; Miya, S.; Sugawara, T. Tetrahedron Lett. 1988, 29,
5673.
(23) Tre´court, F.; Que´guiner, G. J . Chem. Res., Synop. 1982, 76.
(24) Compound 39 has properties different from the previously
reported isomer, 5-oxo-5H-pyrano[2,3-c:6,5-c′]dipyridine: Tre´court, F.;
Marsais, F.; Gu¨ngo¨r, T.; Que´guiner, G. J . Chem. Soc., Perkin Trans. 1
1990, 2409.

Synthesis of N-Confused Porphyrin Analogues
J . Org. Chem., Vol. 61, No. 23, 1996 8129
Matrix isolation was carried out using a 10 cm long, 0.8 cm
diameter quartz tube in an oven directly attached to the
vacuum shroud of a closed cycle liquid He crystat.25 Ar and
N2 were used as matrix media, which were passed over the
sample while it was sublimed and co-condensed as a matrix
at ca. 12 K on a BaF2 window for IR spectroscopy.
Sch em e 7
Photolyses were carried out with a low-pressure Hg lamp
(75 W, 254 nm maximum output; Gra¨ntzel, Karlsruhe, Ger-
many) or a high-pressure Xe-Hg lamp (1000 W; Hanovia).
Matrix FTIR, 1H and 13C NMR, mass spectra (70 eV; direct
insertion), GC-MS, and UV spectra were recorded as previ-
ously indicated.13a Melting points are uncorrected.
Com p u ta tion a l Deta ils. Density functional theory cal-
culations were performed with the Gaussian 92/DFT26 pro-
grams. Geometry optimizations were carried out with the
B-LYP formulation27 of density functional theory, i.e., the
Becke exchange functional (B)27a in combination with the Lee-
Yang-Parr correlation functional (LYP),27b using the 6-31G*
basis set.28 Harmonic vibrational frequencies and infrared
intensities were computed at these equilibrium geometries. A
recent study has shown that unscaled B-LYP/6-31G* frequen-
cies are suitable for prediction of experimental fundamental
frequencies.29
3-Dia zo-4-oxod ih yd r oqu in olin e (14). This compound
was synthesized according to the literature procedure15 with
modifications. The diazonium chloride of 3-diazo-4-oxodihy-
drdquinoline (13) was obtained by diazotization of 3-amino-
4-oxoquinoline (12) in aqueous HCl/EtOH solution with
NaNO2.15 After being dried in vacuo overnight, 13 (690 mg;
3.32 mmol) was mixed with excess Na2CO3 powder in dry
ether. The suspension was stirred in the dark for 3 h before
filtering. The ether solution was evaporated and diazo com-
pound 14 purified by sublimation (460 mg; 2.69 mmol, 81%):
mp 128 °C dec (lit.15 mp 129-130 °C); IR (KBr) 2156 vs, 1627.5
vs, 1606 s, 1585.5 s, 1546 s, 1466.5 m, 1329 w, 1278 m, 1248
m, 1176 m, 1148 w, 864 w, 765 m, 747 w, 684.5 m cm-1; IR
(12K, Ar matrix) 2137 m, 2110 vs, 1662 s, 1611 w, 1580 m,
1567 w, 1549.5 s, 1470 w, 1386 w, 1319 w, 1277 m, 1251.5 s,
1238 w, 1200 w, 1180 m, 1145.5 w, 965 w, 951 w, 906 w, 863
NMR spectroscopy of 34 reveals a symmetrical struc-
ture. In the 1H NMR, the H(C-2) in the pyrrole rings
inside the macrocyle give rise to a low-field signal at 8.78
ppm. In the mass spectrum, ions due to tetramer, trimer,
dimer, and monomer are seen. The structure of tetramer
34 was investigated theoretically by density functional
theory at the B-LYP/6-31G* level. 34 is predicted to have
a symmetrical geometry, with S4 symmetry (Figure 2).
The four inner C-H bonds inside the macrocyclic ring
are in the arrangement of a (not fully regular) tetrahe-
dron (D2d, with H-H distances 2.408 and 2.565 Å; see
Figure 2). The cavity described by the four central
carbons atoms has C-C cross-ring distances of 4.306 Å,
similar to those in normal porphyrins. Finally, we note
that the formation of tetramer 34 from four molecules of
ketene 33 is calculated to be highly exothermic, by 533
kJ mol-1 (BLYP/6-31G*).
1
w, 773 m cm-1; H NMR (CDCl3, 400 Hz) δ 8.65 (s, 1H), 8.27
(dd, J ) 7.79, 1.35 Hz, 1H), 7.81 (dd, J ) 7.26, 0.8 Hz, 1H),
7.77 (td, J ) 6.98, 1.62 Hz, 1H), 7.50 (td, J ) 7.66, 1.36 Hz,
1H); 13C NMR (CDCl3, 400 Hz) δ 164.4, 140.1, 138.0, 127.6,
123.5, 121.5, 118.9, 118.7; UV (Ar matrix, 12 K) λmax 244, 361
nm; MS m/ z 171 (M+; 32), 143 (100), 115 (79), 88 (26), 62 (11).
3-Dia zo-3H-in d ole (18). 3-Oximino-3H-indole (18)16 (5 g,
0.034 mol) was dissolved in 240 mL of degassed H2O together
with 36 mL of 1 N NaOH (0.0030 mol) at 0 °C under N2. Concd
(12.5%) NH3 solution (12 mL) was added with stirring. NaOCl
(60 mL, 5.25%) was then added dropwise over a 15 min period.
Stirring continued for a further 30 min before leaving the
mixture in the refrigerator (4 °C) overnight. The solution was
then extracted with ether, washed with water, and dried.
3-Diazo-3H-indole thus obtained was stable in ether solution
at room temperature for several days: 1H NMR (CDCl3, 200
MHz) δ 8.60 (s, 1H), 7.83-7.87 (d, J ) 6.9 Hz, 1H), 7.50-7.55
(d, J ) 5.9 Hz, 1H), 7.31 (m, 2H); IR (12K, Ar matrix) 2092
vs, 1462 m, 1428.5 s, 1424.5 vs, 1385.5 w, 1271 s, 1199 m,
As ester 37 is easy to synthesize and purify; it is the
ideal precusor for the preparation of tetramer 34.
1113.5 w, 1063.5 w, 846.5 w, 763 w, 759 w cm-1
. This
compound decomposed partially during the measurement of
1
the H NMR spectrum, and a 13C NMR spectrum or elemental
Exp er im en ta l Section
analysis were not obtainable.
Ap p a r a tu s. Preparative FVT was carried out in electrically
heated quartz tubes, either 32 or 40 cm long, of 2 cm diameter.
Samples were sublimed into the horizontal pyrolysis tube
using a sublimation oven. The system was evacuated to ca.
10-5 mbar and continuously pumped during thermolysis using
a turbomolecular pump. The product was trapped at 77 K on
a cold finger or in a U-tube. For trapping of the products with
methanol, the pyrolysate was cocondensed with methanol
vapor on the 77 K cold finger, whereby methanol was intro-
duced between the exit of the oven and the cold finger. In
reactions where removal of methanol was desired, products
were condensed in a U-tube at 77 K, and methanol was
removed by continuous pumping during warmup. Further
details of the FVT apparatus have been published.25
3-Ca r bon yl-3H-in d ole (16). (a ) Dia zo com p ou n d 14 was
matrix isolated in Ar by sublimation (10-5 mbar) at 50 °C onto
(25) Wentrup, C.; Blanch, R.; Briehl, H.; Gross, G. J . Am. Chem.
Soc. 1988, 110, 1874.
(26) Frisch, M. J .; Trucks, G. W.; Schlegel, H. B.; Gill, P. M. W.;
J ohnson, B. G.; Wong, M. W.; Foresman, J . B.; Robb, M. A.; Head-
Gordon, M.; Replogle, E. S.; Gomperts, R.; Andres, J . L.; Raghavachari,
K.; Binkley, J . S.; Gonzalez, C.; Martin, R. L.; Fox, D. J .; DeFrees, D.
J .; Baker, J .; Stewart, J . J . P.; Pople, J . A. GAUSSIAN 92/DFT,
Revision F.2, Gaussian, Inc. Pittsburgh, PA, 1993.
(27) (a) Becke, A. D. Phys. Rev. A 1988, 38, 3098. (b) Lee, C.; Yang,
W.; Parr, R. G. Phys. Rev. B 1988, 37, 785.
(28) Hehre, W. J .; Radom, L.; Schleyer, P. v. R.; Pople, J . A. Ab Initio
Molecular Orbital Theory; Wiley: New York, 1986.
(29) Wong, M. W. Chem. Phys. Lett. 1996, 256, 391.

8130 J . Org. Chem., Vol. 61, No. 23, 1996
Qiao et al.
F igu r e 2. Optimized (B-LYP/6-31G*) geometry of tetramer 34.
a BaF2 window with Ar as the carrier gas. On 254 nm
photolysis for 13 min, 14 was completely converted to 3-car-
bonyl-3H-indole (16): IR (Ar, 12 K) 2139 vs, 1485.5 w, 1473
m, 1448 m, 1331 w, 1284.5 w, 1193.5 w, 1109 w, 1059 m, 868
w, 761 w cm-1; UV (Ar, 12 K) λmax 228, 305.5 nm. Ketene 16
was also matrix isolated by FVT of 14 at 650 °C with Ar as a
carrier gas (10-3 mbar).
(d ) Wa r m u p of Keten e 16. Ketene 16 was matrix isolated
in Ar by photolysis of 14 as described above. The deposition
window was slowly warmed to room temperature and any
reaction was monitored by IR spectroscopy. Ketene 16 disap-
peared at 58 K, and after warming to rt, tetramer 20 was
identified by IR and 1H NMR (DMSO-d6) of the material
remaining on the window. In these experiments, the formation
of tetramer 20 was detected already below 70 K (Figure 5 in
the supporting information).
(b) Dia zo Com p ou n d 18 was sublimed onto a 12 K BaF2
window in vacuo (10-5 mbar) at 70 °C with CO as a carrier
gas. The resulting matrix of 18 in CO matrix was irradiated
at 254 nm for 15 min. Ketene 16 thus formed had IR (CO
3-Dia zo-4-oxop yr id in e (31). This compound was synthe-
sized by a modification of the literature procedure.30 The
diazonium chloride of 3-diazo-4-oxopyridine (30) was obtained
by diazotization at 0 °C of 3-amino-4-oxopyridine (29) in EtOH
solution containing dry HCl with isoamyl nitrite. 30 was dried
under vacuum overnight. A 500 mg sample of 30 was mixed
with excess Na2CO3 powder in dry ether solution. The
suspension was stirred in the dark for 3 h before the solid was
filtered off. The ether solution of 3-diazo-4-oxopyridine (31)
thus formed was stable, but the compound slowly decomposed
in the solid state: IR (Ar, 12 K) 2129.5 vs, 1657 vs, 1574.5 m,
1492.5 s, 1470.5 w, 1381 w, 1234.5 m, 1228 m, 1188.5 m,
1134.5 m, 1118 w, 981 w, 865.5 m, 832 w, 827 w cm-1; IR (N2,
12 K) 2144 vs, 1651 vs, 1570 w, 1489 s, 1474 w, 1383 w, 1237
w, 1228 m, 1188 m, 1135 m, 1120 w, 984 w, 867 m, 830 w, 827
matrix, 12 K) 1484, 1471.5, 1447, 1062 cm-1
.
(c) F r om An h yd r id e 28. The anhydride of indole 3-car-
boxylic acid (28)20 was subjected to FVT at 940 °C, and the
products were matrix isolated in Ar on a 12 K BaF2 window.
Ketene 16 was observed by IR spectroscopy (2136 cm-1),
together with the rearranged ketene 24 (2127 cm-1).
(d ) Tetr a m er 20 was subjected to FVT at 700 °C with
matrix isolation of the product in Ar matrix. By IR spectros-
copy, ketene 16 was the only compound observed: IR (Ar
matrix, 12 K) 2136 vs, 1487 w, 1473.5 w, 1462 w, 1059 w, 867
w, 766 w, 761 w cm-1
.
1,3′-Tetr a k is(in d olylm eth a n on e) (20). (a ) F VT of Di-
a zo Com p ou n d 14. A 900 mg (5.26 mmol) sample of 14 was
sublimed at 85 °C (10-4 mbar) and preparatively thermolyzed
at 500 °C. A light yellow compound condensing on the
glassware immediately outside the oven was collected and
washed with acetone. The insoluble white solid (675 mg, 75
%) was identified spectroscopically as 1,3′-tetrakis(indolyl-
methanone) (20):14 mp 290-292 °C dec (lit.14b mp >300 °C);
1H NMR (DMSO-d6, 400 MHz, 50 °C) δ 7.50-7.53 (t, H5, J )
7 Hz), 7.53-7.57 (t, H6, J ) 6 Hz), 8.16-8.18 (d, H4, J ) 8
Hz), 8.38-8.48 (d, H7 J ) 7.3 Hz), 9.42 (s, H2); 13C NMR
(DMSO-d6, 400 MHz, 50 °C) δ 112.4 (C3), 115.3 (C7), 120.7
(C4), 124.8 (C5), 125.7 (C6), 128.0 (C3a ), 135.6 (C7a), 137.4
(C2), 162.9 (C8).14 The definitive NMR signal assignments
were based on two-dimensional 13C-1H correlations (COLOC)
with both short- and long-range couplings: IR (KBr) 1689 vs,
1539.5 m, 1479.5 w, 1450 vs, 1379.5 m, 1326.5 w, 1263 w, 1200
w cm-1
.
3-Ca r bon yl-3H-p yr r ole (33). (a ) Dia zo com p ou n d 31
was matrix isolated in Ar or N2 by sublimation (10-5 mbar) at
60 K onto a BaF2 window with Ar or N2 as a carrier gas. On
254 nm photolysis for 5 min, 31 was completely converted to
3-carbonyl-3H-pyrrole (33): IR (Ar, 12 K) 2145.5 vs, 1504.5
w, 1460 w, 1444.5 m, 1397 w, 1292.5 m, 1258 w, 1200.5 w,
1097.5 w, 937 m, 825 w cm-1; IR (N2, 12 K) 2150 vs, 1501.5 w,
1445.5 m, 1393.5 w, 1293 m, 1258.5 w, 1202w, 1097 w, 1029
w, 938 m, 862 w cm-1
.
(b) F r om Ester 37. This compound21 was subjected to FVT
at 1100 °C and the product Ar matrix isolated on a 12 K BaF2
window. The isolated products were identified by IR spec-
troscopy as ketene 33 (2145.5 cm-1), CO (2139, 2149 cm-1),
and MeOH (1034, 3732 cm-1).
vs, 1151.5 s, 1019 m, 831.5 s, 748.5 s cm-1 MS m/ z 572 (M+,
1,3′-Tet r a k is(p yr r olylm et h a n on e) (34). (a ) F VT of
Ester 37. A sample of 37 (150 mg; 0.98 mmol) was sublimed
at 120-140 °C and pyrolyzed at 1000 °C, using a U-tube as
cold trap. The eliminated MeOH was trapped in the U-tube
cooled with liq. N2. A light brown compound depositing
immediately outside the oven was collected and washed with
acetone, affording 94 mg (90%) of 1,3′-tetrakis(pyrrolylmetha-
none) (34). 34 can be further purified by sublimation at 260
°C (10-4 mbar): mp 200-202 °C dec; 1H NMR (DMSO-d6, 400
MHz) δ 8.78 (dd, J ) 1.9, 1.9 Hz, 1H), 7.66 (dd, J ) 3.49, 1.91
Hz, 1H), 6.92 (dd, J ) 3.71, 1.59 Hz, 1H); 13C NMR (DMSO-
d6, 400 MHz) δ 161.9, 131.5, 121.35, 118.4, 113.1; IR (KBr)
1711 vs, 1544 s, 1493 s, 1396 m, 1357.5 s, 1316 s, 1158 vs,
;
27), 429 (10), 286 (55), 143 (100), 115 (35), 89 (8).
(b) F VP of An h yd r id e 27. A 50 mg sample of 27, obtained
in quantitative yield by stirring indole-3-carboxylic acid in
acetic anhydride solution, was sublimed at 100-125 °C and
preparatively thermolyzed at 750 °C. The compound isolated
from the cold finger was examed by GC-MS and NMR and
identified as indole (85%). Acetone washing of the residual
material in the sublimation tube afforded 4 mg (∼10%) of
insoluble white solid, which was identified as tetramer 20 by
1H NMR spectroscopy and comparison with the above sample.
(c) F VP of An h yd r id e 28. A 100 mg sample of 28 was
sublimed at ca. 120 °C and subjected to preparative FVT at
700 °C. Indole (30 mg; 77%) was isolated from the cold finger,
and 41 mg (87%) of tetramer 20 was obtained as a white
powder from the residue in the sublimation tube after washing
with acetone.
837 m, 752 m cm-1 MS m/ z 372 (M+, 10.1), 279 (4.9), 186
;
(16.2), 94 (100), 93 (25.6), 73 (14.3), 66 (19.1), 44 (12.1), 39
(30) Su¨s, O.; Mo¨ller, K. J ustus Liebigs Ann. Chem. 1955, 593, 91.

Synthesis of N-Confused Porphyrin Analogues
J . Org. Chem., Vol. 61, No. 23, 1996 8131
12
(25.4), 38 (17.2) cm-1; HRMS calcd for
found 372.0858.
C
H12N4O4 372.0857,
930 m, 795 m cm-1; MS m/ z 198 (M+, 100), 170 (30), 169 (20),
20
143 (22), 122 (15), 115 (25), 53 (25), 50 (30); HRMS calcd for
12
(b) F VT of Dia zo Com p ou n d 31. Diazonium salt 30 (100
mg) was neutralized to 31 with Na2CO3 in ether solution as
described above. The ether was evaporated in the FVT sample
tube, and 31 was subjected to FVT at 550 °C. A small amount
of brown material was isolated from the tube immediately
outside the pyrolysis oven. This compound was identified as
tetramer 34 by MS and 1H NMR (ca. 1% yield).
10-Oxo-10H-p yr a n o[3,2-c:5,6-c′]d ip yr id in e (39). Ester
37 (50 mg; 0.33 mmol) was subjected to preparative FVT at
710 °C, trapping the products in a U-tube at liquid N2
temperature. A powder (8 mg) was isolated from the glass-
ware immediately outside the oven, washed with pentane, and
identified as 4-hydroxynicotinic acid. The compound isolated
from the U-tube was subjected to SiO2 column chromatogra-
phy, eluting with acetone, which afforded 23 mg (61%) of 10-
oxo-10H-pyrano[3,2-c:5,6-c′]dipyridine (39): mp 200-201 °C;
1H NMR (acetone-d6, 400 MHz) δ 9.35 (s, 1H), 8.89 (d, J )
5.82 Hz, 1H), 7.62 (d, J ) 5.82 Hz, 1H); 13C NMR (acetone-d6,
400 MHz) δ 162.05, 156.03, 155.00, 150.26, 138.81, 113.75; IR
(KBr) 1673 vs, 1612 s, 1599 vs, 1468 m, 1344 m, 1170.5 m,
C
H6N2O2 198.0429, found 198.0427. Anal. Calcd for
11
C
11H6N2O2: C, 66.67; H, 3.05; N, 14.14. Found: C, 66.84; H,
3.065; N, 14.07.24
Ack n ow led gm en t. This work was supported by the
Australian Research Council. We thank Dr D. J .
Brecknell for performing the force field calculations.19
Su p p or tin g In for m a tion Ava ila ble: Copies of 1H (400
MHz) and 13C NMR (100 MHz) spectra of 34. Figures 3-7
showing observed matrix IR spectra of 16, 14, 18, 20, 31, and
33, calculated IR spectrum (BLYP/6-31G*) of 16, calculated
IR (MP2/6-31G*), and optimized geometries (BLYP/6-31G*) of
16 and 33 (7 pages). This material is contained in libraries
on microfiche, immediately follows this article in the microfilm
version of the journal, and can be ordered from the ACS; see
any current masthead page for ordering information.
J O9609438