Accelerated Bergman cyclization of porphyrinic-enediynes

Article abstract of DOI:10.1039/b212923j

The Bergman cyclization of simple diethynylporphyrinic-enediynes exhibits a double activation barrier to the formation of Bergman cyclized product. Addition of H-atom acceptor accelerates the formation of the picenoporphyrin, indicating that the second barrier is rate limiting.

Full text of DOI:10.1039/b212923j

Accelerated Bergman cyclization of porphyrinic-enediynes†
Mahendra Nath, John C. Huffman and Jeffrey M. Zaleski*
Department of Chemistry, Indiana University, Bloomington, IN, USA. E-mail: Zaleski@indiana.edu;
Fax: 812-855-8300; Tel: 812-855-2134
Received (in Purdue, IN, USA) 31st December 2002, Accepted 14th February 2003
First published as an Advance Article on the web 5th March 2003
The Bergman cyclization of simple diethynylporphyrinic-
enediynes exhibits a double activation barrier to the forma-
tion of Bergman cyclized product. Addition of H-atom
acceptor accelerates the formation of the picenoporphyrin,
indicating that the second barrier is rate limiting.
The high thermal barrier to the Bergman cyclized picenopor-
phyrin product of 2a in solution reported by Smith et al. (190
°C, 8 h)6 seemed contrary to the general empirical relationship
between DSC temperature and solution-state Bergman cycliza-
tion in simple metalloenediyne structures.2–5,9 Moreover,
heating of 2a,b in MeOH/CHCl3 over a wide temperature range
(25–80 °C) gives very sluggish reactivity with small ( < 5%)
and nearly constant product yield. These factors suggested that
cycloaromatization may not be the high barrier step in
picenoporphyrin product formation. Rather, loss of H2 (or 2H·)
could be inhibiting product formation.
Our interests in activation of enediyne frameworks toward
Bergman cyclization1 and the formation of the reactive
1,4-phenyl diradical intermediate2–4 led us to probe the
syntheses and reactivities of conjugated enediyne constructs.5
Recently, Smith et al. showed that conjugated porphyrin-
enediynes could be prepared in good yields and would
thermally cyclize to generate substituted Ni-picenoporphyrin
derivatives.6 The proposed mechanism for this unusual self-
quenching reaction involves a tetrahydro intermediate that must
lose 2 moles of H2 to promote aromatization to the Ni-
picenoporphyrin product. The atypical mechanistic inter-
mediate, coupled with the quantitative recovery of the –TMS
starting material upon thermal activation, further prompted our
interests in whether the activation barrier to product formation
was governed in part by the loss of H2, as well as the usual
contributions to Bergman cyclization such as steric strain in the
transition states7 and products, as well as the alkyne termini
separation.8,9
Addition of 2 equiv of the 2e2/2H+ acceptor DDQ (DDQ =
2,3-dichloro-5,6-dicyano-1,4-benzoquinone) to 2a,b at room
temperature in CHCl3/MeOH (2+1) yields picenoporphyrin
products 3a,b in 30–40% yield within 30 min (Scheme 2) with
no additional heating.
The green products 3a6 and 3b, are straightforwardly
detected in solution by 1H NMR. For 3b, a diagnostic resonance
from the two protons of the new Bergman cyclized ring appear
at d = 8.93, while those of the modified meso-phenyl rings are
detected as distinct doublets at d = 8.91 and 9.56 ppm, a triplet
at 7.91 ppm, as well as a convoluted triplet at 7.76 ppm. In
addition, the electronic absorption spectrum of 3b (CH2Cl2) is
markedly red-shifted from that of 2b exhibiting a Soret band
with lmax = 459 nm, and four distinct Q-bands at lmax = 600,
622, 649, and 684 nm, thereby confirming the increase in p-
To address these issues, we have prepared Ni(II) and free base
2,3-diethynyl-5,10,15,20-tetraphenylporphyrins with –TMS or
–H at the alkyne termini positions† using the method reported
by Smith et al. (Scheme 1).6
Reaction of the Ni(II) or free base (2H) 2,3-dibromo-
5,10,15,20-tetraphenylporphyrin10 (1a,b) with Me3Sn–·–TMS
(TMS = trimethylsilyl) over a Pd(0) catalyst,11 and subsequent
deprotection with base under aqueous conditions yields the
2,3-diethynyl-5,10,15,20-tetraphenylporphyrin, 2a,b.
The X-ray structures of 2a,b (Fig. 1) show that the Ni(II
)
derivative is highly distorted and adopts a mixed ruffled/saddled
conformation,12 while the free base analogue is relatively
planar, exhibiting only a modest mixed ruffle/saddle distortion
which removes the coplanarity of the alkyne termini carbons.
These conformations lead to room temperature stable structures
with solid-state Bergman cyclization temperatures of 172 and
160 °C, respectively, as measured by differential scanning
calorimetry.
Fig. 1 X-Ray structures of 2a,b. Alkyne termini separation 2a: C27…C29
= 4.44 Å; 2b: C26…C28 = 4.09 Å. Thermal ellipsoids are illustrated at
50% probability.
Scheme 1 Synthesis of diethynylporphyrinic-enediyne 2a,b.
† Electronic supplementary information (ESI) available: syntheses, charac-
terization of 2a–4b, crystallographic files (CCDC 200680–200685) in CIF
format. See http://www.rsc.org/suppdata/cc/b2/b212923j/
Scheme 2 Bergman cyclization of 2a,b in the presence of DDQ to generate
picenoporphyrins 3a,b.
858
CHEM. COMMUN., 2003, 858–859
This journal is © The Royal Society of Chemistry 2003

conjugation in 3b. Formation of picenoporphyrin products 3a,b
have also been confirmed by X-ray crystallography (Fig. 2).
The structure of 3a6 exhibits a planar piceno unit within a
strongly ruffled porphyrin backbone. In contrast, the structure
of 3b is nearly planar, with only a modest ruffle distortion.
subsequently must lose 2 moles of H2 (or 4 H·) to rearomatize
to the picenoporphyrin product. Since under our solution
conditions (CHCl3/MeOH) in the absence of DDQ, only
starting material and product are observed by NMR up to 85 °C,
a more highly reactive intermediate than the tetrahydro species
(e.g. dihydrodiradical, III, Scheme 3) must be in equilibrium
with starting material in the absence of DDQ. It is clear that the
H-atom donor ability of the solution may be influencing the
intermediate en route to picenoporphyrin product.
Fig. 2 X-ray structures of 3a,b. Thermal ellipsoids are illustrated at 50%
probability.
The –TMS derivatives (4a,b) can also be prepared by
reaction 1a or 1b with (trimethylsilylacetylene)trimethyl-
stannane as that shown for the preparation of 2a prior to
deprotection.6 The X-ray structure of 4a once again exhibits a
mixed ruffled/saddled distortion, while 4b is generally planar
with only a modest out-of-plane projection of the alkynes from
the ring in a syn-configuration (Fig. 3). The alkyne termini
separations are very similar, indicating that the conformation of
the porphyrin has a modest effect, in these cases, on the
disposition of the enediyne unit.
Scheme 3 Illustration of the influence of DDQ on the proposed porphyrinic-
enediyne-picenoporphyrin reaction coordinate.
Heating of 4a,b in the solid-state reveals strong exothermic
peaks in the DSC traces at 332 and 388 °C. For 4a, this feature
is immediately preceded by a melting endotherm, indicating a
change of state and negating the ability to compare these
temperatures directly.
Reaction of 4a in CHCl3/MeOH in the presence of DDQ at 25
°C for 24 h reveals a slow decomposition of the starting material
without detectable formation of picenoporphyrin product.
These results are in contrast to those observed upon reaction of
2a,b with DDQ under the same conditions. This comparison,
coupled with the high DSC temperatures for 4a,b, and the
inability to obtain picenoporphyrin product in solution at
accessible temperatures from 4a alone,6 reflect an increased
activation barrier to Bergman cyclization for 4a,b due at least in
part to the steric crowding of the –TMS functionalities in the
product.
The ability to generate Bergman cyclized product in 30–40%
yield under ambient conditions in the presence of DDQ suggests
that the primary reaction coordinate for picenoporphyrin
product formation can be divided into at least two sequential
steps with two significant activation barriers: 1) a Bergman
cyclization and radical addition reaction, and 2) a hydrogen
loss/rearomatization step (Scheme 3). In light of the poor yield
of product in the absence of DDQ, the Bergman cyclization/
addition step must be facile, but contain a rapid equilibrium
between the proposed intermediate and starting material as the
tetrahydro species is not detected under our conditions. This
also suggests that the second step is product limiting, and that
the activation barrier for loss of hydrogen is greater than that for
the Bergman cyclization/addition step. From a biological
perspective, the ability of external substrates or reactants to
drive Bergman cyclization reactions to completion suggests that
environmental factors may be important for modulating ene-
diyne reactivity under varying solution conditions.
The observation that DDQ greatly enhances generation of
3a,b without detection of an intermediate is in agreement with,
but not identical to the mechanism proposed by Smith et al.
involving formation of the quasi-stable tetrahydro intermediate
via H-atom donation by 1,4-cyclohexadiene.6 This species
Notes and references
1 R. G. Bergman, Acc. Chem. Res., 1973, 6, 25.
2 P. J. Benites, D. S. Rawat and J. M. Zaleski, J. Am. Chem. Soc., 2000,
122, 7208.
3 D. S. Rawat and J. M. Zaleski, J. Am. Chem. Soc., 2001, 123, 9675.
4 T. Chandra, M. Pink and J. M. Zaleski, Inorg. Chem., 2001, 40, 5878.
5 B. J. Kraft, N. L. Coalter, M. Nath, A. E. Clark, A. R. Siedle, J. C.
Huffman and J. M. Zaleski, Inorg. Chem., 2003, 42(5), 1663.
6 H. Aihara, L. Jaquinod, D. J. Nurco and K. M. Smith, Angew. Chem.,
Int. Ed., 2001, 40, 3439.
7 J. P. Snyder, J. Am. Chem. Soc., 1990, 112, 5367.
8 K. C. Nicolaou, G. Zuccarello, C. Riemer, V. A. Estevez and W.-M. Dai,
J. Am. Chem. Soc., 1992, 114, 7360.
9 B. P. Warner, S. P. Millar, R. D. Broene and S. L. Buchwald, Science,
1995, 269, 814.
10 L. K. Jaquinod, R. G. Shea and K. M. Smith, Tetrahedron, 1999, 55,
13151.
11 S. G. DiMagno, V. Victor, S.-Y. Lin and M. J. Therien, J. Am. Chem.
Soc., 1993, 115, 2513.
12 C. J. Medforth, M. O. Senge, K. M. Smith, L. D. Sparks and J. A.
Shelnutt, J. Am. Chem. Soc., 1992, 114, 9859.
Fig. 3 X-ray structures of 4a,b. Alkyne termini separation 4a: C27…C33 =
4.19 Å; 4b: C26…C32 = 4.16 Å. Thermal ellipsoids are illustrated at 50%
probability.
CHEM. COMMUN., 2003, 858–859
859