Synthesis and reactions of phosphaporphyrins: Reconstruction of π-skeleton triggered by oxygenation of a core phosphorus atom

Article abstract of DOI:10.1021/jo902060b

(Chemical Equation Presented) The synthesis, structures, optical and redox properties, and reactivity of phosphaporphyrins are reported. The 21-phosphaporphyrin (P,N₃-porphyrin) and 23-phospha-21-thiaporphyrin (P,S, N₂-porphyrin) wer

Full text of DOI:10.1021/jo902060b

pubs.acs.org/joc
Synthesis and Reactions of Phosphaporphyrins: Reconstruction of
π-Skeleton Triggered by Oxygenation of a Core Phosphorus Atom
Takashi Nakabuchi,† Makoto Nakashima,† Shinya Fujishige,‡ Haruyuki Nakano,‡
Yoshihiro Matano,*,† and Hiroshi Imahori†,§,^
†Department of Molecular Engineering, Graduate School of Engineering, Kyoto University,
Nishikyo-ku, Kyoto 615-8510, Japan, ‡Department of Chemistry, Graduate School of Sciences,
Kyushu University, Fukuoka 812-8581, Japan, §Institute for Integrated Cell-Material Sciences (iCeMS),
Kyoto University, Nishikyo-ku, Kyoto 615-8510, Japan, and ^Fukui Institute for Fundamental Chemistry,
Kyoto University, Sakyo-ku, Kyoto 606-8103, Japan
matano@scl.kyoto-u.ac.jp
Received October 7, 2009
The synthesis, structures, optical and redox properties, and reactivity of phosphaporphyrins are
reported. The 21-phosphaporphyrin (P,N3-porphyrin) and 23-phospha-21-thiaporphyrin (P,S,
N2-porphyrin) were prepared via acid-promoted dehydrative condensation between a phosphatri-
pyrrane and the corresponding 2,5-bis[hydroxy(phenyl)methyl]heteroles followed by 2,3-dichloro-
5,6-dicyanobenzoquinone oxidation. Experimental (NMR, UV-vis, and X-ray analyses) and
theoretical (DFT calculations) results suggest that the 18π aromaticity inherent in regular
N4-porphyrins was maintained in these phosphaporphyrins. X-ray crystallography revealed a
slightly distorted 18π aromatic ring for the P,N3-porphyrin with the phosphole and three pyrrole
rings tilted from the 24-atoms mean plane by 9.6ꢀ and 3.8-15.4ꢀ, respectively. DFT calculations on
model compounds showed that the P,X,N2-porphyrins (X = N, S) possess considerably small
HOMO-LUMO gaps as compared with N4- and S,N3-porphyrins, which is reflected in the red-
shifted absorptions, low oxidation potentials, and high reduction potentials of the phosphaporphyr-
ins. The P-oxygenation of the P,X,N2-porphyrins with H2O2 has been found to lead to the formation
of different types of products. The 18π P,N3-porphyrin was transformed into the 22π aromatic P(O),
N3-porphyrin accompanied by the π extension at the peripheral C3 bridge, whereas the 18π P,S,N2-
porphyrin was converted to the isophlorin-type 20π antiaromatic P(O),S,N2-porphyrin. In both of
the reactions, simple P-oxygenated 18π P(O),X,N2-porphyrins were formed as the initial products,
which were subsequently transformed into the 22π or 20π porphyrins. The two reaction courses from
18π to 20π/22π are apparently determined by the combination of the core heteroatoms (i.e., P,N3 or
P,S,N2) and the structure of the peripherally fused carbocycles. The present results demonstrate that
the incorporation of a phosphorus atom into the core is not only a highly promising way to modify the
fundamental properties of the porphyrin 18π system but also a reliable tool to stabilize uncommon
22π and 20π systems through the chemical modifications at the core phosphorus atom.
Introduction
and materials chemistry. To perturb optical, electrochemi-
cal, and coordinating properties of porphyrins, chemical
modifications of the macrocyclic platform have generally
been employed. Core modification, namely, the replacement
Porphyrins are one of the most widely studied macrocyclic
heterocycles because of their important roles in biological
DOI: 10.1021/jo902060b
r
Published on Web 12/14/2009
J. Org. Chem. 2010, 75, 375–389 375
2009 American Chemical Society

Nakabuchi et al.
JOCArticle
of one or more core nitrogen atoms by other heteroatoms or
carbon, is a highly promising chemical modification meth-
odology.1-5 Recent extensive studies on core-modified por-
phyrins have disclosed that the electronic structures of their
π circuits differ significantly from those of regular porphyr-
ins (hereafter denoted as the N4-porphyrins). For instance,
Soret and Q bands of 21-chalcogena- and 21,23-dichalco-
genaporphyrins are red-shifted from those of N4-porphyrins,
and the bathochromic shifts of these heteroporphyrins are
strongly dependent on the relevant core chalcogens (O, S, Se,
Te).2f The incorporation of four chalcogens (O, S) at the core
has been successfully utilized to stabilize planar 20π systems
(isophlorin skeletons) that are difficult to construct with
N4-porphyrins.6 An additional promising aspect of the
core-modified porphyrins is the unique reactivity endowed
by the core elements. For example, the 21-tellurapoprhyrins
undergo the Te-O exchange reaction7 or oxidative chlor-
ination8 at the tellurium center via Te-oxygenation, and
carbaporphyrinoids, typically in their metal complex forms,
exhibit a variety of reactivities such as alkylation,9 cyaniza-
tion,10 diphenylphosphanylation,11 halogenation,12 nitra-
tion,12 oxygenation,13 pyridination,14 and internal fusion.12
Despite these encouraging findings on the core-modified
porphyrins, the types of elements introduced into the core
have been limited to carbon, silicon,15 and chalcogens.
Under these circumstances, we were interested in the chem-
istry of phosphaporphyrin, a heavy analogue of the parent
N4-porphyrin, because the phosphole subunit provides un-
precedented optical, electrochemical, and coordinating
properties to the porphyrin platform.
In sharp contrast to pyrrole, phosphole forms a trigonal
pyramid structure at the phosphorus center and behaves
basically as a neutral P ligand and as a highly conjugative cis-
1,3-dienic π system.16 Such structural and electronic proper-
ties of phosphole are beneficial for the construction of an
unprecedented class of core-modified porphyrins. Most im-
portantly, the tricoordinate (σ3) phosphorus center can be
converted to various tetracoordinate (σ4) forms by simple
chemical modification such as P-oxygenation (to σ4-PdO),
P-thioxygenation (to σ4-PdS), and P-metal coordination (to
σ4-P-metal) with its active lone electron pair.17 This implies
that the electronic structure of a phosphaporphyrin π system
is readily tunable by the introduction of P substituents.
In 2003, Delaere and Nguyen predicted the electronic
structures and optical properties of unsubstituted 21-phos-
pha- and 21,23-diphosphaporphyrins based on density
functional theory (DFT) calculations and concluded that
these phosphaporphyrins would exhibit reasonable aroma-
ticity.18 However, the synthesis of phosphaporphyrins has
not been reported until recently, presumably due to the lack
of their potential precursors. The reactivity of phosphole
is well-known to arise from the low aromatic character of the
five-membered ring and the high nucleophilicity of the
σ3-phosphorus atom. Accordingly, conventional methods
(Friedel-Crafts alkylation and direct lithiation) that have
ꢀ
(1) For reviews, see: (a) Latos-Graz_ynski, L. In The Porphyrin Handbook;
Kadish, K. M., Smith, K. M., Guilard, R., Eds.; Academic Press: San Diego, 2000;
Vol 2, Chapter 14. (b) Furuta, H.; Maeda, H.; Osuka, A. Chem. Commun. 2002,
1795–1804. (c) Srinivasan, A.; Furuta, H. Acc. Chem. Res. 2005, 38, 10–20. (d)
Chandrashekar, T. K.; Venkatraman, S. Acc. Chem. Res. 2003, 36, 676–691. (e)
ꢀ
ꢀ
Stepien, M.; Latos-Graz_ynski, L. Acc. Chem. Res. 2005, 38, 88–98. (f)
ꢀ
Chmielewski, P. J.; Latos-Graz_ynski, L. Coord. Chem. Rev. 2005, 249, 2510–
2533. (g) Gupta, I.; Ravikanth, M. Coord. Chem. Rev. 2006, 250, 468–518. (h)
Lash, T. D. Eur. J. Org. Chem. 2007, 5461–5481. (i) Misra, R.; Chandrashekar,
T. K. Acc. Chem. Res. 2008, 41, 265–279.
(2) Chalcogen-containing porphyrins. For example, see: (a) Broadhurst,
M. J.; Grigg, R.; Johnson, A. W. J. Chem. Soc. C 1971, 3681–3690. (b)
Ulman, A.; Manassen, J. J. Am. Chem. Soc. 1975, 97, 6540–6544. (c)
_ ꢀ
Latos-Grazynski, L.; Lisowski, J.; Olmstead, M. M.; Balch, A. L. J. Am.
_ ꢀ
Chem. Soc. 1987, 109, 4428–4429. (d) Chmielewski, P. J.; Latos-Grazynski,
L.; Olmstead, M. M.; Blalch, A. L. Chem.-Eur. J. 1997, 3, 268–278. (e)
Sridevi, B.; Narayanan, S. J.; Srinivasan, A.; Chandrashekar, T. K.;
Subramanian, J. J. Chem. Soc., Dalton Trans. 1998, 1979–1984. (f) Abe,
M.; Hilmey, D. G.; Stilts, C. E.; Sukumaran, D. K.; Detty, M. R. Organo-
metallics 2002, 21, 2986–2992. (g) Agarwal, N.; Ravikanth, M. Tetrahedron
2004, 60, 4739–4747. (h) Uno, H.; Shimizu, Y.; Uoyama, H.; Tanaka, Y.;
Okujima, T.; Ono, N. Eur. J. Org. Chem. 2008, 87–98. (i) Cho, D.-G.; Plitt,
P.; Lim, S. K.; Lynch, V.; Hong, S,-J.; Lee, C.-H.; Sessler, J. L. J. Am. Chem.
Soc. 2008, 130, 10502–10503.
(3) Carbaporphyrinoids. For example, see: (a) Berlin, K.; Breitmaier, E.
Angew. Chem., Int. Ed. Engl. 1994, 33, 1246–1247. (b) Berlin, K. Angew.
Chem., Int. Ed. Engl. 1996, 35, 1820–1822. (c) Lash, T. D.; Hayes, M. J.
(10) Xiao, Z.; Patrick, B. O.; Dolphin, D. Chem. Commun. 2003, 1062–
1063.
_ ꢀ
(11) Grzegorzek, N.; Pawlicki, M.; Szterenberg, L.; Latos-Grazynski, L.
ꢀ
Angew. Chem., Int. Ed. Engl. 1997, 36, 840–842. (d) Stepien, M.; Latos-
_ ꢀ
J. Am. Chem. Soc. 2009, 131, 7224–7225.
(12) Furuta, H.; Ishizuka, T.; Osuka, A.; Ogawa, T. J. Am. Chem. Soc.
2000, 122, 5748–5757.
Grazynski, L. Chem.-Eur. J. 2001, 7, 5113–5117. (e) Lash, T. D.; Hayes, M.
J.; Spence, J. D.; Muckey, M. A.; Ferrence, G. M.; Szczepura, L. F. J. Org.
_ ꢀ
Chem. 2002, 67, 4860–4874. (f) Sprutta, N.; Swiderska, M.; Latos-Grazynski,
(13) (a) Hung, C.-H.; Chen, W.-C.; Lee, G.-H.; Peng, S.-M. Chem. Commun.
2002, 1516–1517. (b) Xiao, Z.; Patrick, B. O.; Dolphin, D. Inorg. Chem. 2003, 42,
L. J. Am. Chem. Soc. 2005, 127, 13108–13109.
(4) N-Confused porphyrins. For example, see: (a) Furuta, H.; Asano, T.;
Ogawa, T. J. Am. Chem. Soc. 1994, 116, 767–768. (b) Chmielewski, P. J.;
ꢀ
_ ꢀ
8125–8127. (c) Stepien, M.; Latos-Grazynski, L.; Szterenberg, L. J. Org. Chem.
2007, 72, 2259–2270. (d) Lash, T. D.; Szymanski, J. T.; Ferrence, G. M. J. Org.
Chem. 2007, 72, 6481–6492.
_ ꢀ
Latos-Grazynski, L.; Rachlewicz, K.; Glowiak, T. Angew. Chem., Int. Ed.
ꢀ
(14) Stepien, M.; Latos-Grazynski, L. Org. Lett. 2003, 5, 3379–3381.
(15) Latos-Grazynski and co-workers attempted to synthesize 21-sila-
porphyrin, which led them to the serendipitous discovery of a unique
transformation of 21-silaphlorin into a carbacorrole. Skonieczny, J.;
_ ꢀ
Engl. 1994, 33, 779–781. (c) Furuta, H.; Maeda, H.; Osuka, A. J. Am. Chem.
Soc. 2000, 122, 803–807. (d) Furuta, H.; Ishizuka, T.; Osuka, A. J. Am. Chem.
Soc. 2002, 124, 5622–5623. (e) Morimoto, T.; Uno, H.; Furuta, H. Angew.
Chem., Int. Ed. 2007, 46, 3672–3675. (f) Toganoh, M.; Kimura, T.; Furuta,
H. Chem.-Eur. J. 2008, 14, 10585–10594.
_ ꢀ
_ ꢀ
Latos-Grazynski, L.; Szterenberg, L. Chem.-Eur. J. 2008, 14, 4861–4874.
(5) Pyriporphyrins. For example, see: (a) Berlin, K.; Breitmaier, E.
Angew. Chem., Int. Ed. Engl. 1994, 33, 219–220. (b) Lash, T. D.; Chaney,
(16) For reviews, see: (a) Mathey, F. Chem. Rev. 1988, 88, 429–453. (b)
Quin, L. D. In Comprehensive Heterocyclic Chemistry II; Katritzky, A. R., Rees,
C. W., Scriven, E. F. V., Eds.; Elsevier: Oxford, 1996; Chapter 15. (c) Hissler, M.;
_ ꢀ
S. T. Chem.-Eur. J. 1996, 2, 944–948. (c) Mysliborski, R.; Latos-Grazynski,
ꢀ
L.; Szterenberg, L. Eur. J. Org. Chem. 2006, 3064–3068. (d) Sim, E.-K.;
Jeong, S.-D.; Yoon, D.-W.; Hong, S.-J.; Kang, Y.; Lee, C.-H. Org. Lett.
2006, 8, 3355–3358. (e) Lash, T. D.; Pokharel, K.; Serling, J. M.; Yant, V. R.;
Ferrence, G. M. Org. Lett. 2007, 9, 2863–2866.
Dyer, P. W.; Reau, R. Coord. Chem. Rev. 2003, 244, 1–44. (d) Mathey, F. Angew.
ꢀ
Chem., Int. Ed. 2003, 42, 1578–1604. (e) Baumgartner, T.; Reau, R. Chem. Rev.
2006, 106, 4681-4727; (correction) 2007, 107, 303. (f) Hobbs, M. G.; Baum-
ꢀ
gartner, T. Eur. J. Inorg. Chem. 2007, 23, 3611–3628. (g) Reau, R.; Dyer, P. W. In
(6) Chalcogen-containing isophlorins: (a) Pohl, M.; Schmickler, H.; Lex,
J.; Vogel, E. Angew. Chem., Int. Ed. Engl. 1991, 30, 1693–1697. (b) Reddy,
J. S.; Anand, V. G. J. Am. Chem. Soc. 2008, 130, 3718–3719.
Comprehensive Heterocyclic Chemistry III; Ramsden, C. A., Scriven, E. F. V.,
Taylor, R. J. K., Eds.; Elsevier: Oxford, 2008; Chapter 3.15. (h) Matano, Y.;
Imahori, H. Org. Biomol. Chem. 2009, 7, 1258–1271.
_ ꢀ
(7) Latos-Grazynski, L.; Pacholska, E.; Chmielewski, P. J.; Olmstead,
(17) For example, see: (a) Hay, C.; Hissler, M.; Fischmeister, C.;
ꢀ
ꢀ
M. M.; Blalch, A. L. Angew. Chem., Int. Ed. Engl. 1995, 34, 2252–2254.
(8) Abe, M.; Detty, M. R.; Gerlits, O. O.; Sukumaran, D. K. Organome-
tallics 2004, 23, 4513–4518.
Rault-Berthelot, J.; Toupet, L.; Nyulaszi, L.; Reau, R. Chem.-Eur. J.
2001, 7, 4222–4236. (b) Delaere, D.; Nguyen, M. T.; Vanquickenborne, L.
G. J. Phys. Chem. A 2003, 107, 838–846. (c) Baumgartner, T.; Bergmans, W.;
_ ꢀ
ꢀ
ꢀ
ꢀ
(9) (a) Chmielewski, P. J.; Latos-Grazynski, L.; Glowiak, T. J. Am. Chem.
Karpati, T.; Neumann, T.; Nieger, M.; Nyulaszi, L. Chem.-Eur. J. 2005, 11,
4687–4699.
(18) Delaere, D.; Nguyen, M. T. Chem. Phys. Lett. 2003, 376, 329–337.
Soc. 1996, 118, 5690–5701. (b) Schmidt, I.; Chmielewski, P. J. Inorg. Chem.
2003, 42, 5579–5593.
376 J. Org. Chem. Vol. 75, No. 2, 2010

Nakabuchi et al.
JOCArticle
CHART 1. (a) 21-Phospha- and 21,23-Diphosphaporphyrins
Calculated by Delaere and Nguyen. (b) Mathey’s P-Confused
Carbaporphyrinoid
CHART 2. Phosphaporphyrins 1N and 1S and Thiaporphyrins
2N and 2S
derivatives. We envisioned that these comparative studies
would highlight the characteristics of phosphorus in por-
phyrin chemistry and heteroatom chemistry.
been widely used for the R-functionalization of pyrrole and
thiophene are not applicable to phosphole. Recently, Math-
ey’s group and our group independently reported convenient
methods for the synthesis of 2,5-difunctionalized phospholes
based on sequential [1,5]-sigmatropic shifts in P-functiona-
lized phospholes19 and TiII-mediated cyclization of difunc-
tionalized diynes,20 respectively. 2,5-Bis[hydroxymethyl]-
type phospholes, suitable starting materials of phosphapor-
phyrins, are now readily available in gram scale.
In 2007, Mathey and co-workers reported an attempt to
prepare a phosphaporphyrin via acid-promoted [3 þ 1]
condensation of a 2,5-bis[phenyl(hydroxy)methyl]phosphole,
where the phosphorus atom was masked as the σ4-PdS form,
with a tripyrrane. However, this approach resulted in the
formation of a small amount of “P-confused” carbaporphyr-
inoid (Chart 1).21,22 In 2006, we reported the first synthesis of
a phosphaporphyrin of the P,S,N2-type. This synthesis was
achieved by using a different [3 þ 1] condensation approach
starting from a phosphatripyrrane (vide infra),23a and this
method was later applied to the synthesis of a P,N3-type
phosphaporphyrin.23b With these encouraging results avail-
able, we aimed to uncover the structures, aromaticity,
and optical/electrochemical properties of phosphaporphyr-
ins by comparison with those of known porphyrins. We also
aimed to clarify the reactivities of phosphaporphyrins, espe-
cially those at the core phosphorus atom, as well as the
fundamental properties of the resulting P-functionalized
In this report, full details of the synthesis and reactions of
21-phosphaporphyrin (P,N3-porphyrin) 1N and 23-phos-
pha-21-thiaporphyrin (P,S,N2-porphyrin) 1S (Chart 2) are
described.23,24 The structures, aromaticity, and optical/elec-
trochemical properties of 1N, 1S, and their P-functionalized
derivatives are discussed in detail. The replacement of the
N-H unit of the N4-porphyrin with the P-Ph unit induced
dramatic alternations in the fundamental properties and
reactivity of the porphyrin ring. In particular, these 18π
phosphaporphyrins underwent unique π-reconstructions
triggered by P-oxygenation, resulting in the formation of
22π or 20π systems depending on the combination of the core
heteroatoms and the structure of the peripherally fused
carbocycles.25
Results and Discussion
1. Synthesis and Characterization of Phosphaporphyr-
ins. 1.1. Synthesis. Scheme 1 illustrates the synthesis of 21-
phosphaporphyrin 1N and 23-phospha-21-thiaporphyrin
1S. The BF3-promoted dehydrative condensation of
P-masked phosphatripyrrane 323a with 2,5-bis[hydroxy-
(phenyl)methyl]-pyrrole 4N,26 -thiophene 4S,2b or -furan
4O2d afforded P-masked P,X,N2-porphyrinogens 5X-S
(X = N, S, O) in 9-43% yields as a mixture of two or three
diastereomers. Desulfurization of 5X-S with excess
P(NMe2)3 in refluxing toluene produced the corresponding
σ3-P,X,N2-porphyrinogens 5X (X = N, S, O) in good yields.
Compounds 5X were subsequently treated with 3.3 equiv of
2,3-dichloro-5,6-dicyanobenzoquinone (DDQ) at room
temperature. In the DDQ oxidation of 5N, two major
products were obtained. Flash column chromatography of
the crude reaction mixture on alumina (CH2Cl2/hexane = 1/1)
followed by recrystallization from MeOH gave a reddish
purple solid (Rf = 0.9) and a dark-purple solid (Rf = 0.5) in
17% and 8% yields, respectively. On the basis of the spectro-
scopic and crystallographic analyses (vide infra), the less
polar compound was characterized as the expected 21-phos-
phaporphyrin 1N, whereas the more polar compound
was found to be the P-oxo-type phosphaporphyrin 6-O.
Under similar conditions, 5S reacted with DDQ to afford
two major products. Flash column chromatography of the
(19) (a) Holand, S.; Jeanjean, M.; Mathey, F. Angew. Chem., Int. Ed.
Engl. 1997, 36, 98–100. (b) Toullec, P; Mathey, F. Synlett 2001, 1977–1979.
(20) Matano, Y.; Miyajima, T.; Nakabuchi, T.; Matsutani, Y.; Imahori,
H. J. Org. Chem. 2006, 71, 5792–5795.
(21) Duan, Z.; Clochard, M.; Donnadieu, B.; Mathey, F.; Tham, F. S.
Organometallics 2007, 26, 3617–3620.
(22) Mathey and co-workers prepared some phosphole-containing
macrocycles: (a) Laporte, F.; Mercier, F.; Ricard, L.; Mathey, F. J. Am.
Chem. Soc. 1994, 116, 3306–3311. (b) Deschamps, E.; Ricard, L.; Mathey, F.
J. Chem. Soc., Chem. Commun. 1995, 1561. (c) Mercier, F.; Laporte, F.;
€
Ricard, L.; Mathey, F.; Schroder, M.; Regitz, M. Angew. Chem., Int. Ed.
Engl. 1997, 36, 2364–2366.
(23) For preliminary results, see: (a) Matano, Y.; Nakabuchi, T.;
Miyajima, T.; Imahori, H.; Nakano, H. Org. Lett. 2006, 8, 5713–5716. (b)
Matano, Y.; Nakashima, M.; Nakabuchi, T.; Imahori, H.; Fujishige, S.;
Nakano, H. Org. Lett. 2008, 10, 553–556.
(24) Recently we reported phosphole-containing calixpyrroles and calix-
phyrins: (a) Matano, Y.; Nakabuchi, T.; Miyajima, T.; Imahori, H.
Organometallics 2006, 25, 3105–3107. (b) Matano, Y.; Miyajima, T.; Naka-
buchi, T.; Imahori, H.; Ochi, N.; Sakaki, S. J. Am. Chem. Soc. 2006, 128,
11760–11761. (c) Matano, Y.; Miyajima, T.; Ochi, N.; Nakabuchi, T.; Shiro, M.;
Nakao, Y.; Sakaki, S.; Imahori, H. J. Am. Chem. Soc. 2008, 130, 990-1002;
(correction) 2009, 131, 14123. (d) Nakabuchi, T.; Matano, Y.; Imahori, H.
Organometallics 2008, 27, 3142–3152. (e) Matano, Y.; Imahori, H. Acc. Chem.
Res. 2009, 42, 1193–1204. (f) Matano, Y.; Fujita, M.; Miyajima, T.; Imahori, H.
Phosphorus, Sulfur Silicon Relat. Elem. In press. (g) Matano, Y.; Fujita, M.;
Miyajima, T.; Imahori, H. Organometallics 2009, 28, 6213–6217.
(25) Recently, we found that 1S undergoes redox-coupled complexation
with zero-valent group 10 metals (M = Ni, Pd, Pt) to afford air-stable MII-P,
S,N2-isophlorin complexes: Matano, Y.; Nakabuchi, T.; Fujishige, S.;
Nakano, H.; Imahori, H. J. Am. Chem. Soc. 2008, 130, 16446–16447.
(26) Heo, P.-Y.; Lee, C.-H. Bull. Korean Chem. Soc. 1996, 17, 515–520.
J. Org. Chem. Vol. 75, No. 2, 2010 377

Nakabuchi et al.
JOCArticle
SCHEME 1. Synthesis of P,X,N2-Porphyrins
SCHEME 2. Synthesis of P,N3-Porphyrin Bearing a Peripheral C4-Bridge
crude reaction mixture on alumina (CH2Cl2/hexane = 1/1)
followed by recrystallization from MeOH gave reddish
purple (Rf = 0.7) and dark-green (Rf = 0.9) solids in 5%
and 1% yields, respectively. In this case, the more polar
compound was found to be the expected 23-phospha-21-
thiaporphyrin 1S, and the less polar compound was the
P-oxo derivative 7-O.27 Although the isolated yields of the
products were low, the formation of 1X, 6-O, and 7-O was
quite reproducible. Unfortunately, all attempts to prepare
21-phospha-23-oxaporphyrin 1O from 5O were unsuccess-
ful; DDQ oxidation of 5O resulted in the formation of a
complex mixture. Another P,N3-porphyrin, 1N0, which bears
a peripheral C4-bridge at the phosphole ring, was also
prepared from the corresponding precursors 30 and 4N
(Scheme 2). In this case, the P-oxo byproducts like 6-O
were not obtained. All of the phosphaporphyrins 1N, 1N0,
and 1S were observed to be stable in the solid state but slowly
decomposed in solution under aerobic conditions and in
room light. We assume that 6-O and 7-O were mainly
formed by the P-oxidation of the corresponding P,X,N2-
porphyrins 1N and 1S during the reaction conditions em-
ployed. The details of these reactions will be discussed later.
To compare core-modification effects on the fundamental
properties of porphyrin π systems, 21-thia- and 21,23-dithia-
porphyrins, 2N and 2S25 (Chart 2), were newly prepared by a
similar [3 þ 1]-type condensation of thiatripyrrane S2 with
4X (X = N, S) followed by DDQ oxidation (Scheme S1 in
Supporting Information). In the synthesis of 2X, no S-oxo
(sulfoxide or sulfone) type byproducts were produced.
1.2. Characterization. The diagnostic spectral features
of P,X,N2-porphyrins 1N, 1N0, and 1S are as follows. In
the high-resolution FAB-MS (HR-FAB-MS) spectra, the
parent ion peaks (Mþ) were observed at m/z 595.2182 for
1N (calcd for C41H30N3P, 595.2177), 609.2335 for 1N0 (calcd
for C42H32N3P, 609.2334), and 612.1788 for 1S (calcd
for C41H29N2PS, 612.1789). In the 1H NMR spectrum
of 1N, meso and pyrrole-β protons appeared at δ 10.18
(27) After the preliminary results were published (ref 23a), we noticed that
a small amount of 7-O was also isolable beside 1S.
378 J. Org. Chem. Vol. 75, No. 2, 2010

Nakabuchi et al.
JOCArticle
FIGURE 1. 1H NMR spectra of phosphaporphyrins (a) 1N, (b) 1S, (c) 6-O, (d) 6-S, and (e) 7-O in CDCl3. Asterisks (/) indicate the peaks of
residual solvents.
3
(d, JP,H = 16.1 Hz) and 8.35-8.67 ppm, respectively,
whereas the NH and P-phenyl protons appeared at δ
-0.59 and 2.43-5.68 ppm, respectively (Figure 1a; Table
S1 in Supporting Information). It is apparent that the
significant downfield and upfield positions of these peaks
stem from ring current effects of the porphyrin 18π-electron
The data for the heterole-β protons indicate that the ring
current effects increase in the order S2,N2-porphyrin > P,S,
N2-porphyrin ≈ S,N3-porphyrin > P,N3-porphyrin. The
diminished ring current effects in the P,X,N2-porphyrins as
compared with the corresponding S,X,N2-porphyrins may
be ascribed to the slightly deviated π-planes in the phospha-
porphyrins (vide infra). The upfield shifts of the P-phenyl
protons of 1X versus 8X increase in the order ortho > meta
> para, implying that the P-phenyl group in 1X stands
above the porphyrin π-plane. As summarized in Table S1
in Supporting Information, the ring current effects
also emerged as upfield positions of the 31P peaks of 1N
(δ -5.2 ppm), 1N0 (δ -32.6 ppm), and 1S (δ 18.6 ppm)
relative to the 31P peaks of the respective precursors, 5N (δ
30.4-32.7 ppm), 5N0 (δ 2.6-4.1 ppm), and 5S (δ 32.7-33.5
ppm).29 The structural and spectral properties of the P-oxo
porphyrins 6-O and 7-O will be summarized later.
1
circuit. In the H NMR spectra of 1N0 and 1S (Figure 1b),
similar diatropic ring current effects were observed. To
evaluate and compare the ring current effects quantitatively,
the difference in chemical shifts (Δδ) of the heterole-β and P-
phenyl protons between 1X (X = N, S) and calixphyrins 8X
(X = N, S)24b,c (Chart 3) was used as an index. As the π
circuits in 8X are interrupted at the sp3-hybridized meso
carbons, the positive/negative signs and the absolute values
of Δδ (Table 1) represent the diatropicity (paratropicity) and
the degree of the aromaticity (antiaromaticity) of the π-
circuits of 1X. The downfield shifts of the heterole-β protons
observed for the S,X,N2-porphyrins 2X (X = N, S) versus
the S,X,N2-calixphyrins 9X (X = N, S)28 are also summar-
ized in Table 1.
(29) Phosphorus-31 chemical shifts of phospholes are sensitive to the
substituents on the phosphorus atom and β-carbons. Since we have no data
about 31P chemical shifts of P-oxo calixphyrins and the C4-bridge type
calixphyrins, the 31P chemical shifts of the Cn-bridge type P(E),X,N2-
porphyrins (n = 3, 4; E = lone pair, O, S; X = N, S) are compared with
those of the corresponding porphyrinogens.
(28) Matano, Y.; Miyajima, T.; Ochi, N.; Nakao, Y.; Sakaki, S.; Imahori,
H. J. Org. Chem. 2008, 73, 5139–5142.
J. Org. Chem. Vol. 75, No. 2, 2010 379

Nakabuchi et al.
JOCArticle
a
˚
CHART 3. P,X,N2-Calixphyrins 8X, 8X-S, and 14 and S,X,
N2-Calixphyrins 9X (X = N, S)
TABLE 2. Selected Bond Lengths and Distances (A) in the Crystal
Structures
compound
TABLE 1. Differences in the 1H NMR Chemical Shifts (Δδ, ppm)
between Porphyrins and Calixphyrinsa
1N; 18π
E, X, N = lone
pair, NH, N
6-O; 22π
E, X, N = O,
N, NH
7-O; 20π
E, X, N = O,
S, NH
a
1.38
1.42
1.37
1.41
1.45
1.34
1.46
1.37
1.36
1.40
1.43
1.42
1.35
1.39
1.54
1.50
4.56
3.45
1.48
1.42
1.37
1.40
1.42
1.37
1.42
1.38
1.37
1.40
1.41
1.44
1.35
1.38
1.39
1.40
4.47
4.37
2.65
1.36
1.48
1.36
1.44
1.39
1.40
1.39
1.38
1.38
1.47
1.38
1.44
1.34
1.77
1.54
1.50
4.69
4.33
2.73
b (b0)
c (c0)
d (d0)
e (e0)
f (f0)
g (g0)
h (h0)
i (i0)
j (j0)
k (k0)
l (l0)
m
heterole-β
(β1, β2, β3)
P-Ph
(ortho, meta, para)
porphyrin/calixphyrin
n (n0)
o (o0)
p (p0)
1N/8N
þ2.08, þ1.79, þ2.70
þ1.93, þ1.65, þ2.53
þ2.29, þ2.07, þ2.62
þ2.29, þ1.97, þ2.89
þ2.35, þ2.13, þ3.10
þ1.56, þ1.48, þ1.98
þ0.90, þ0.97, þ1.42
-0.35, -0.97, -0.76
-5.1, -2.0, -1.6
-5.0, -2.1, -2.0
-5.3, -2.0, -1.6
1N0/8N
1S/8S
2N/9N
N22 N24
3 3 3
P
X
2S/9S
3 3 3
b
PdO N22/24
3 3 3
6-O/8N-S
6-S/8N-S
7-O/14
-2.5, -0.9, -0.6
þ0.1, þ0.1, þ0.1
þ0.7, þ0.5, þ0.6
aAverage values between y and y0 (y denotes b-l and n-p). bAverage
values.
aMeasured in CDCl3. Chemical shifts for the calixphyrin references
were reported in ref 24b, 24c (for 8X, 8X-S, and 14) and ref 28 (for 9X).
Δδ = δ (porphyrin) - δ (calixphyrin). The plus sign indicates a down-
field shift and the minus sign indicates an upfield shift.
1X and S,X,N2-porphyrins 2X (X = N, S) in CH2Cl2 are
displayed in Figure 2.30 The Soret bands of 1N and 1S appeared
at λab 431 and 440 nm (Table 3), respectively, which are
red-shifted compared with those of 2N (λab 418 nm), 2S (λab
426 nm), and 5,10,15,20-tetraphenylporphyrin (H2TPP, 10)
(λab 411 nm).2f,31 The longest Q-type transitions of 1N (λab
698 nm) and 1S (λab 718 nm) were also observed at longer
wavelengths relative to those of 2N (λab 660 nm), 2S (λab 682 nm),
and 10 (λab 646 nm). Thus, the replacement of an N-H unit
with a P-Ph unit shifts the Soret and Q bands to the longer
wavelengths more strongly than does the replacement with a
sulfur atom. The UV-vis absorption spectrum of 1N0 (Soret, λab
431 nm; Q, λab 692 nm) was quite similar to the spectrum
recorded on 1N. The S,X,N2-porphyrins 2X were found to be
fluorescent, whereas the P,X,N2-porphyrins 1X were observed to
be nonfluorescent.
The structure of 1N was unambiguously determined by
X-ray crystallography.23b The selected bond lengths and
distances are summarized in Table 2. The porphyrin ring in
1N is slightly distorted, wherein the phosphole and three
pyrrole rings are tilted from the 24-atom mean plane with
dihedral angles of 9.6ꢀ and 3.8-15.4ꢀ, respectively. The
differences between the contiguous carbon(meso)-carbon-
(R) bond lengths (i.e., the absolute values of d-c and k-j in
˚
Table 2) are notably suppressed (0.03-0.04 A), which
indicates the efficient π-electron delocalization through
the 18π circuit of 1N. In this regard, 1N possesses
aromatic character in terms of the geometrical criterion.
The phosphorus center adopts a trigonal pyramidal geometry
with C-P-C bond angles of 91.2(13)-101.99(12)ꢀ
P
(
C-P-C = 293.7ꢀ), and the P-phenyl group is located above
(30) In the previous reports (ref 23), we underestimated the molar
absorbance coefficients (ε) for 1N and 1S owing to their gradual decomposi-
tion during sample preparation. Therefore, the samples for the UV-vis
measurement of 1X were prepared quickly under very weak light and
remeasured. We checked that the spectra did not change under the measure-
ment conditions.
(31) Delaere and Nguyen performed TD-DFT calculations on P,N3-
porphyrin 1N(H)-m at the B3LYP/SV(P) level and predicted the red-shifts
of the Soret and Q bands by 18-23 and 16-26 nm, respectively, as compared
to those of N4-porphyrin 10-m (see ref 18).
the porphyrin π-plane as suggested by the NMR observations.
These structural features imply that the porphyrin π-circuit
does not involve the lone electron pair on the phosphorus. The
core of 1N provides a trapezoid cavity with N22-N24 and
˚
P-N23 distances of 4.56 and 3.45 A, respectively.
2. Optical and Electrochemical Properties. 2.1. Optical
Property. The UV-vis absorption spectra of P,X,N2-porphyrins
380 J. Org. Chem. Vol. 75, No. 2, 2010

Nakabuchi et al.
JOCArticle
positive Ered,1 than H2TPP 10 (Eox,1 = þ0.58 V, Ered,1
=
-1.73 V, ΔE = 2.31 V).2f
3. DFT Calculations. To gain more insight into the struc-
ture, aromaticity, and optical and electrochemical properties
of P,X,N2-porphyrins 1X and S,X,N2-porphyrins 2X,
DFT calculations on a series of model compounds 1N-m,
1S-m, 1N(H)-m,32 2N-m, 10-m, and 11-m (Chart 4) at the
B3LYP/6-311G(d,p) level were carried out. The optimized
structures are shown in Figure 4, and selected bond lengths
and distances are summarized in Table S2 in Supporting
Information. The differences between the contiguous
carbon(meso)-carbon(R) bond lengths (i.e., the absolute
˚
values of d-c and k-j in Table S2) of 1N-m (0.01-0.03 A)
˚
and 1S-m (0.01-0.02 A) are comparable to those of 10-m
(0.01 A). In the structures of P,X,N2-models 1X-m, each
˚
P
phosphorus atom forms a pyramidal geometry (
=
C-P-C
299.5ꢀ for both 1N-m and 1S-m), and the P-phenyl group is
located above the macrocyclic plane (Figure 4a and b). These
theoretical results are in good agreement with the experi-
mental results (NMR and X-ray) observed for 1N and 1S.
In contrast to the completely planar S,N3-model 2N-m, P,
N3-models 1N-m and 1N(H)-m possess slightly distorted
π planes. The porphyrin π plane of 1S-m deviates more than
that of 1N-m, probably as a result of the electrostatic
repulsion between the two relatively large heteroatoms, P
and S, making a dish-like macrocyclic platform with a
slightly deeper bottom. The comparison of the optimized
structure of the P-phenyl model 1N-m with that of the P-H
model 1N(H)-m confirmed that the phenyl group attached to
the phosphorus causes a slightly larger distortion of the
macrocyclic framework (Figure 4a versus c; dihedral angles
between the phosphole ring and the 24-atom mean plane
for 1N-m and 1N(H)-m are 12.1ꢀ and 2.7ꢀ, respectively).
However, the deviation caused by the introduction of the
phenyl group is considerably smaller than that observed for
N4-porphyrins; the π planes of the N-alkyl and -aryl por-
phyrins have been reported to deviate substantially from
planarity.33 The N-phenyl pyrrole ring of 11-m is tilted from
the 24-atom mean plane with a dihedral angle of 34.9ꢀ, and
FIGURE 2. UV-vis absorption spectra of (a) 1N (red) and 2N
(black) and (b) 1S (red) and 2S (black) in CH2Cl2.
TABLE 3. Absorption and Fluorescence Properties in CH2Cl2
λ
ab, nm (ε, 103 M-1 cm-1
)
compound
Soret band
431 (177)
431 (172)
440 (164)
418 (218)
426 (188)
Q-bandb
λ
em, nma (ΦF, %)
1N
698 (3.1)
nfc
1N0
1S
692 (3.2)
718 (3.1)
660 (3.2)
682 (3.4)
nfc
nfc
2N
2S
665, 735 (0.94)d
685, 760 (0.46)e
6-O
6-S
7-O
422 (42.1), 494 (77.2) 964 (2.5)
424 (25.7), 485 (29.4) 750-1200 (sh)
394 (66.1)
1130 (0.8)
aExcited at 506 nm for 2N and 2S. bThe longest absorption maxima.
cNonfluorescent. dFluorescence quantum yield relative to ΦF of 2S.
eAbsolute fluorescence quantum yield determined by a calibrated inte-
grating sphere system.
the phenyl-bound nitrogen atom takes a flattened pyramidal
geometry with the sum of the bond angles of
342.9ꢀ (Figure 4f). This is because the core nitrogen atom
in 11-m aptly retains the sp2 hybridization, even when it is
phenylated.
P
=
C-N-C
2.2. Electrochemical Property. Redox potentials of 1X
were measured by cyclic voltammetry (CV) and/or differen-
tial pulse voltammetry (DPV) and are compared with the
potentials of 2X and 10 (Figure 3; Figure S1 and Table S4 in
Supporting Information). The electrochemical oxidation
processes of 1N and 1S were found to be irreversible, and
the first oxidation potentials (Eox,1) of 1N and 1S, deter-
mined by DPV, were þ0.38 and þ0.45 V (vs Fc/Fcþ),
respectively, both of which are more cathodic than Eox,1 of
2N (þ0.62 V) and 2S (þ0.70 V). On the other hand, the
electrochemical reduction of 1N and 1S occurred reversibly
Nucleus-independent chemical shifts (NICS) have been
used as indices for evaluating the aromaticity of the porphyrin
rings.34 The NICS values at the center of the four heteroa-
toms of 1N-m and 1S-m were determined as -13.8 and -15.7
ppm, respectively (Table 4). The absolute NICS values of
1N-m and 1S-m are somewhat smaller than those of the N4-
porphyrin 10-m (-16.5 ppm) and S,N3-porphyrin 2N-m
(32) In ref 18, Delaere and Nguyen calculated the optimized structure and
orbital energies of 1N(H)-m at the B3LYP/6-31G* and B3LYP/SV(P) levels,
respectively. The orbital energies reported by these authors are as follows:
-5.88 (HOMO-2), -5.50 (HOMO-1), -5.47 (HOMO), -2.69 (LUMO),
and -2.44 eV (LUMOþ1).
or quasi-reversibly, and the first reduction potentials (Ered,1
)
of 1N and 1S were -1.51 and -1.36 V, respectively, which
are more anodic than those of 2N (-1.56 V) and 2S (-1.43
V). As a result, the differences in the redox potentials (ΔE =
Eox,1 - Ered,1) of 1N (1.89 V) and 1S (1.81 V) are significantly
smaller than those of 2N (2.18 V) and 2S (2.13 V). The redox
behavior of 1N0 (Eox,1 = þ0.41 V, Ered,1 = -1.52 V, ΔE =
1.93 V) is quite similar to that of 1N. All of the P,X,
N2-porphyrins 1X showed more negative Eox,1 and more
(33) (a) Lavallee, D. K.; Anderson, O. P. J. Am. Chem. Soc. 1982, 104,
4707–4708. (b) Kuila, D.; Lavallee, D. K. J. Am. Chem. Soc. 1984, 106, 448–
450. (c) Aizawa, S.; Tsuda, Y.; Ito, Y.; Hatano, K.; Funahashi, S. Inorg.
Chem. 1993, 32, 1119–1123.
~
(34) For example, see: (a) Cyranski, M. K.; Krygowski, T. M.;
Wisiorowski, M.; van Eikema Hommes, N. J. R.; Schleyer, P. von R. Angew.
Chem., Int. Ed. 1998, 37, 177–180. (b) Furuta, H.; Maeda, H.; Osuka, A. J.
Org. Chem. 2001, 66, 8563–8572.
J. Org. Chem. Vol. 75, No. 2, 2010 381

Nakabuchi et al.
JOCArticle
FIGURE 3. First oxidation (Eox,1) and reduction (Ered,1) potentials for (a) 1N, (b) 1N0, (c) 1S, (d) 2N, (e) 2S, (f) 6-O, (g) 6-S, and (h) 7-O.
narrower by 0.08 eV than that of 10-m. The DFT calcula-
tions on the P-H model 1N(H)-m32 showed that the replace-
ment of an N-H group with a P-H group stabilizes the
HOMO and LUMO by 0.17 and 0.29 eV, respectively, and
the HOMO-LUMO gap of 1N(H)-m becomes narrower by
0.12 eV than that of 10-m. Therefore, the core modification
with phosphorus has more significant impacts on the frontier
orbitals than that with sulfur.
CHART 4. Model Compounds
Phosphaporphyrins contain three high-energy HOMOs
due to the involvement of a lone electron pair and a sub-
stituent (Ph, H) at the phosphorus atom. In the P-Ph
porphyrin 1N-m, the lone electron pair orbital at the phos-
phorus effectively mixes into one of the two inherent
HOMOs of porphyrin (the one that has large orbital coeffi-
cients at the core and meso positions), thus giving rise to
significantly destabilized HOMO and HOMO-2 (by 0.17 and
0.33 eV) in comparison with those of 1N(H)-m. Presumably,
the σ-π antibonding interaction between the P-C(Ph) σ
orbital and the porphyrin π orbital in 1N-m leads to the
destabilization of these HOMOs. Similar orbital interactions
appear in the P,S,N2-porphyrin 1S-m. The large contribu-
tion of the phosphorus lone pair to the HOMO explains the
unique reactivity of 1X (vide infra). As seen in Figure 5f, the
HOMOs of 11-m are destabilized by 0.07-0.14 eV relative to
those of the N-H counterpart 10-m, which is attributable to
a steric factor rather than an orbital interaction.
(-16.2 ppm) but are sufficiently large enough to maintain
aromaticity. On the basis of these theoretical results,
together with the above-mentioned experimental data, we
can conclude that the inherent aromaticity is maintained in
the P,X,N2-porphyrin 18π systems.
The frontier orbitals of the model compounds are depicted
in Figure 5. The replacement of a core nitrogen atom of
porphyrin with a chalcogen atom is well-known to stabilize
LUMOs more largely than HOMOs. This is commonly
explained by the inductive or electron-withdrawing effect
that the heteroatom has on the frontier orbitals.1a,g Indeed,
in our calculations, HOMO and one of the nearly degener-
ated LUMOs of the S,N3-model 2N-m are stabilized by 0.10
and 0.18 eV, respectively, relative to those of the N4-model
10-m. As a result, the HOMO-LUMO gap of 2N-m becomes
Overall, the HOMO levels of 1N-m (-5.42 eV) and 1S-m
(-5.39 eV) are comparable to the HOMO level of the N4-
porphyrin 10-m (-5.42 eV), whereas the LUMO levels of 1N-
m (-2.78 eV) and 1S-m (-2.90 eV) are largely stabilized by
0.28 and 0.40 eV relative to that of 10-m (-2.50 eV). As a
consequence, the HOMO-LUMO gaps of 1N-m (2.64 eV)
and 1S-m (2.49 eV) are considerably smaller than the gap
calculated for 10-m (2.92 eV). These DFT results are in good
agreement with the above-mentioned experimental results in
that the core modification of porphyrin from the N-H unit or
382 J. Org. Chem. Vol. 75, No. 2, 2010

Nakabuchi et al.
JOCArticle
FIGURE 4. Top (upper) and side (lower) views of the optimized structures for (a) 1N-m, (b) 1S-m, (c) 1N(H)-m, (d) 2N-m, (e) 10-m, and (f) 11-
m. The C, H, N, P, and S atoms are indicated as gray, light blue, blue, orange, and yellow balls, respectively.
TABLE 4. NICS Values for Model Compounds
1N0 was consumed completely, and P-oxo P,N3-porphyrin
1N0-O35 was formed in 94% NMR yield with a small
amount (6%) of its tautomer 120. The structures of 1N0-O
and 120 were characterized on the basis of only the NMR
spectra because of their instability in solution. (For detailed
1H NMR peak assignments of the intermediates presented in
this section, see Figures S4 and S5 in Supporting
31
Information.) The P resonances of 1N0-O (δp 21.8 ppm)
and 120 (δp 39.2 ppm) appeared at the significantly downfield
region in the spectra when compared with the 31P resonance
of 1N0 (δp -32.6 ppm), supporting the notion that the
phosphorus atom was oxygenated. The tautomer 120 shows
an asymmetric 1H peak pattern with two NH protons (δ 13.8;
12.4), two meso protons (δ 5.60, d, 3JP,H = 25.8 Hz; δ 7.12, d,
3JP,H = 35.6 Hz), and an olefinic proton of the peripheral
cyclohexene ring (δ 6.09, pseudo t, J = 4.0 Hz). Most of
1N0-O was converted into 120 within 60 min,36 whereas 120
slowly decomposed to afford a complex mixture after 3 h.
Next, the reaction of the C3-bridge type P,N3-porphyrin 1N
with H2O2 was monitored (Figure 6b). Within 5 min, 1N (δp
-5.2 ppm) was quantitatively converted to P-oxo P,N3-
porphyrin 1N-O (δp 36.4 ppm), which was then tautomer-
ized into 12 (δp 48.2 ppm). The spectral features of 1N-O
and 12 are very close to those of 1N0-O and 120. In contrast
to the reaction of 1N0, however, the mixture of 1N-O and 12
was ultimately (after 3 h) transformed into 6-O, which was
isolated in 77% yield. Finally, the reaction of the P,S,N2-
porphyrin 1S with H2O2 was monitored (Figure 6c). After 10
min, 1S (δp 18.6 ppm) was consumed to generate two types of
P-oxo porphyrins 1S-O (δp 43.8 ppm) and 7-O in 76% and
17% NMR yields, respectively. On standing the solution for
48 h, 1S-O decomposed completely, while 7-O was ob-
tained as the sole isolated product in 16% yield. The above
results imply that the P-oxidation of 1N and 1S is one of the
most plausible pathways leading to 6-O and 7-O in
Scheme 1.
compound
NICS (ppm)
Type A (18π)
1N-m (X = PPh, Y = NH)
1S-m (X = PPh, Y = S)
2N-m (X = S, Y = NH)
10-m (X,Y = NH)
-13.8a
-15.7a
-16.2a
-16.5a
Type B (22π)
6-O-m (E = O)
6-S-m (E = S)
-9.9b
-8.8b
Type C (20π)
7-O-m
2.4b
aCalculated at the center of the four heteroatoms. bCalculated at the
center of the two facing nitrogen atoms.
the S atom to the P-Ph unit basically narrows the HOMO-
LUMO gap.
4. P-Functionalized Phosphaporphyrins. 4.1. Chemical
Functionalization of a Core Phosphorus Atom. The electronic
structures of phosphole-containing π-conjugated systems
are well-known to be effectively altered by the chemical
functionalization of the phosphorus atom.17 We anticipated
that oxidation of the core phosphorus atom would produce
new π systems with different electronic structures. More-
over, the formation of byproducts 6-O and 7-O suggested
to us that the P-oxygenation pathway would be involved in
the synthesis of 1X. With these concepts in mind, we exam-
ined the reactions of 1N, 1N0, and 1S with H2O2 (Scheme 3).
Initially, the reaction of the C4-bridge type P,N3-porphyrin
1N0 with H2O2 in CDCl3 was monitored by 1H NMR
spectroscopy at several intervals (Figure 6a). After 5 min,
From the viewpoint of the syntheses of 6-O and 7-O, it
seemed convenient to oxidize the macrocyclic rings of 5N-O
and 5S-O, which were readily available from the corre-
sponding σ3- P,X,N2-porphyrinogens 5X in high yields (for
details, see Supporting Information). In this context, the
reactions of 5X-O with DDQ were examined (Scheme 4). To
(35) The peripheral and P-phenyl protons of 1X-O (X = N, N0, S) are
upfield-shifted and downfield-shifted, respectively, as compared with those
of 1X (X = N, N0, S), suggesting that 1X-O form more deviated structures
than 1X. See, Table S5 in Supporting Information.
(36) This type of tautomerization from σ4-phospholes to σ4-phospholenes was
ꢀ
reported for the C4-bridge type 2,5-diarylphosphole-P-sulfides: Nyulaszi, L.;
Holloczki, O.; Lescop, C.; Hissler, M.; Reau, R. Org. Biomol. Chem. 2006, 4,
ꢀ
ꢀ
996–998.
J. Org. Chem. Vol. 75, No. 2, 2010 383

Nakabuchi et al.
JOCArticle
FIGURE 5. HOMOs and LUMOs of (a) 1S-m (b) 1N-m, (c) 1N(H)-m, (d) 2N-m, (e) 10-m, and (f) 11-m.
understand the substituent effects on the reactivity, DDQ
oxidations of 5X-S were also examined. Treatment of the P,
N3-porphyrinogens, 5N-O and 5N-S, with DDQ afforded
6-O and 6-S as the sole isolable products in 26% and 17%
yields, respectively (Scheme 4). Hence, 5N-O is also likely to
be a precursor of 6-O in the DDQ oxidation of 5N depicted
in Scheme 1. The overall yield of the conversion from 5N
to 6-O via 5N-O (25%) was better than the yield via 1N
(13%). The P-thioxo substituent in 5N-S appreciably
retarded the efficiency of the ring oxidation; however, the
first PdS phosphaporphyrin 6-S could be isolated as a dark
brown solid. On the other hand, the DDQ oxidations
of 5S-O and 5S-S gave complex mixtures, and the expected
20π porphyrins 7-O and 7-S were not obtained.37
As discussed below, both of the isolated P-oxo porphyrins
6-O and 7-O no longer possess ordinary 18π aromatic
structures; the π system of 6-O is extended to an unprece-
dented type of aromatic 22π system through the peripherally
fused C3-bridge, whereas 7-O possesses an isophlorin-type
antiaromatic 20π structure. Despite containing the isophlorin-
type 20π structure, 7-O is stable in air.38 The formation of
different types of π systems, 6-O and 7-O, can be reason-
ably explained by considering the following factors: (i) the
steric congestion at the core, (ii) the electrostatic and hydro-
gen-bonding interactions among the core components, and
(iii) the redox properties of the macrocyclic π systems
(Scheme 5). Since the P-oxidation increases the steric con-
gestion at the core, factor (i) is basically important for all the
reactions in Scheme 3. Consequently, it is likely that factors
(ii) and (iii) mainly determine the reaction pathways. In the
simple P-oxygenated 18π porphyrins 1N-O and 1S-O,
there must be electrostatic repulsion between the PdO
moiety and the two adjacent nitrogen lone pairs at the core.
By contrast, in 6-O and 7-O, there are two hydrogen bonds
between the NH protons and the PdO moiety (vide infra).
The reconstruction of the 18π structure of 1X-O into the
22π (6-O) or 20π (7-O) structure arises to apparently
relieve the “stress” and gain the attractive hydrogen-bonding
interaction at the core. The intermediate 12, in which the π
circuit is interrupted at the peripheral five-membered ring,
undergoes 2e-oxidation to gain the aromatic stabilization,
giving 6-O as the final product.39 In contrast, in the case
of 1S-O, the formation of a tautomer like 13 is inhibited
probably because such an intermediate is not stabilized
(39) On the basis of the time course of the reaction (Figures S8a and S9 in
Supporting Information) and the result of an independent experiment
(Figures S8b and S10 in Supporting Information), we presume that a coupled
2e-oxidation/2e-reduction between 12 and 1N-O takes place to produce
equimolar amounts of 6-O (oxidized form) and an unidentified intermediate
(reduced form; denoted as IM in Figures S8 and S9). Compound IM (not
shown in Figure 6b) was slowly oxidized in situ to afford 6-O (Scheme S4 in
Supporting Information). For details, see Supporting Information.
(37) DDQ oxidation of the P,O,N2-type porphyrinogen-P-sulfide 5O-S
gave an unexpected N-C fused macrocyclic product (S5). For details, see
Supporting Information.
(38) Isophlorins have an intrinsic nature to undergo 2e-oxidation
(aromatization to 18π structure) under ambient conditions. See ref 6.
384 J. Org. Chem. Vol. 75, No. 2, 2010

Nakabuchi et al.
JOCArticle
SCHEME 3. Reactions of P,X,N2-Porphyrins 1N0, 1N, and 1S with H2O2
sufficiently by one hydrogen bond at the core (Scheme 3).
Instead, the 2e-reduction of 1S-O proceeds rapidly to afford
the 20π porphyrin 7-O,40,41 and the overall reaction
from 1S to 7-O is formally regarded as a hydration. It should
be emphasized that the core phosphorus atom of 1N and 1S
acts as a kind of switch, which realizes a dramatic alteration of
the π-electronic structure by a simple chemical modification.42
4.2. Structures, Aromaticity, and Optical and Electroche-
mical Properties of P-Functionalized Phosphaporphyrins.
Single crystals of 6-O and 7-O were grown from
CH2Cl2-MeOH at room temperature (for 6-O)23b or from
CH2Cl2 at -10 ꢀC (for 7-O). Top and side views of 7-O are
depicted in Figure 7, and selected bond lengths and dis-
tances of 6-O and 7-O are summarized in Table 2. Un-
fortunately, we could not obtain high-quality single crystals
of 6-S. In the crystal structures of 6-O and 7-O, each core
phosphorus atom adopts a tetrahedral geometry with
C-P-C/C-P-O angles of 94.3(3)-107.5(2)ꢀ/108.1(3)-
119.09(18)ꢀ and 94.3(3)-107.5(2)ꢀ/108.1(3)-119.09(18)ꢀ,
respectively. To avoid the steric congestion at the core, the
phosphole rings lean substantially outward. As a conse-
quence, the phosphorus atom deviates from the 24-atom
˚
˚
mean plane by 1.20 A (for 6-O) and 1.22 A (for 7-O), and
the P-C(Ph) bond axis tilts against the mean plane by 114.5ꢀ
(for 6-O) and 115.1ꢀ(for 7-O). These values are larger than
˚
those observed for P,N3-porphyrin 1N (0.34 A and 82.5ꢀ),
indicating that the π circuits in the P-oxo derivatives 6-O
and 7-O are distorted considerably as compared with the
18π circuit of 1N. The 22π aromaticity of 6-O is well
corroborated by the significantly diminished car-
bon-carbon bond length alternation through the 22π-cir-
cuit including the exocyclic C3-bridge. The differences
between the contiguous carbon(meso)-carbon(R) and
carbon(C3-bridge)-carbon(C3-bridge) bond lengths of 6-O
(i.e., the absolute values of d-c and k-j, and o-o0 in Table 2)
˚
are 0.00-0.03 and 0.00 A, respectively. In contrast, not-
able bond length alternation is observed for 7-O, in
which the differences between the contiguous carbon-
˚
(meso)-carbon(R) bond lengths are 0.07-0.09 A. This
indicates that the aromaticity is lost for the 20π-circuit
of 7-O (Table 2). The relatively short PdO N22/24 dis-
3 3 3
˚
˚
tances (2.65 A for 6-O and 2.73 A for 7-O) clearly display
the presence of intramolecular hydrogen-bonds between
the P-oxo group and the two adjacent NH protons.43 This
interaction is likely to play a crucial role in stabilizing the
novel 22π and 20π systems.
(40) Deducing from the reduction potentials of 1S (Ered,1 = -1.36 V,
Ered,2 = -1.56 V), 1N (-1.51 V, -1.74 V), and 1N0 (-1.52 V, -1.72 V),
1S-O should have higher electron affinity than 1N-O and 1N0-O.
(41) On the basis of the time course of the reaction (Figures S8d and S11 in
Supporting Information) and the result of an independent experiment
(Figures S8e and S12 in Supporting Information), we presume that unreacted
1S reduces 1S-O to 7-O and is transformed to 1S-O (Scheme S5 in
Supporting Information). Indeed, the addition of 1S to a solution containing
1S-O produces 7-O quantitatively. For details, see Supporting Informa-
tion.
(43) The N O distances for the NH O hydrogen bonds in dimethyl-
3 3 3
3 3 3
ammonium hydrogen diphenyldiphosphonate were reported to be 2.75-2.77
A: Courtney, B. H.; Juma, B. W. O.; Watkins, S. F.; Fronczek, F. R.;
Stanley, G. G. Acta Crystallogr. 2006, C62, o268–o270.
˚
(42) The reference S,N3-porphyrin 2N does not react with H2O2 even
under more harsh conditions (ca. 100 equiv of H2O2, 24 h).
J. Org. Chem. Vol. 75, No. 2, 2010 385

Nakabuchi et al.
JOCArticle
Figure 8, and selected bond lengths, distances, and angles are
listed in Table S2 in Supporting Information. The bond
parameters of the optimized structures of 6-O-m and
7-O-m are very close to those of the crystal structures of
6-O and 7-O. As shown in Figure 8, the phosphole ring of
6-S-m is tilted more than the phosphole rings of 6-O-m
and 7-O-m to accommodate the bulky PdS moiety in the
core.44 The PdS N22/24 distances (each 3.13 A) calculated
˚
3 3 3
for 6-S-m suggest the presence of intramolecular hydrogen-
bonding interaction between the P-thioxo group and the two
adjacent NH protons.45
In the HR-FAB-MS spectra, 6-O, 6-S, and 7-O clearly
show parent ion peaks (Mþ) at m/z 609.1975 (calcd for
C41H28N3OP, 609.1975 for 6-O), 625.1742 (calcd for
C41H28N3PS, 625.1738 for 6-S), and 630.1895 (calcd for
1
C41H31N2OPS, 630.1895 for 7-O). In the H NMR spec-
trum of 6-O, meso, pyrrole-β, NH, and P-phenyl protons
appear at δ 9.21 (d, 3JP,H = 35.6 Hz), 7.94-8.19, 5.40, and
5.60-6.92 ppm, respectively (Figure 1c). Additionally, the
peripheral protons that originate from a C3-bridge of the
phosphole backbone are observed at δ 8.16 (d, J = 3.9 Hz,
5
2H) and 9.06 (dt, J = 3.9 Hz, JP,H = 5.4 Hz, 1H) ppm,
indicating that the π-circuit goes through the five-membered
fused ring in 6-O. The 31P peak of 6-O appears at δ 28.6
ppm, which is considerably shielded when compared with
that of the P-oxo P,N3-porphyrinogen 5N-O (δ 59.2-59.6
ppm, Table S1 in Supporting Information). These NMR
data obviously reflect diatropic ring-current effects induced
by the aromatic 22π circuit of 6-O; however, both shielding
and deshielding effects are somewhat diminished in compar-
ison with 1N. This reduction in shielding and deshielding is
1
due to the deviated structure of 6-O. The H NMR spec-
trum of 6-S (Figure 1d) is quite different from that of 6-O.
The C3-bridge protons appear at δ 3.80 (d, J = 3.4 Hz, 2H)
and 5.70 (dt, J = 3.4 Hz, JP,H = 5.4 Hz, 1H) ppm,
5
respectively, which are significantly upfield-shifted by 4.36
and 3.36 ppm from those of 6-O. In contrast, the P-phenyl
protons of 6-S (δ 7.5-8.23 ppm) are downfield-shifted by
0.6-2.63 ppm from those of 6-O. As mentioned above, the
phosphole ring of 6-S (6-S-m) is considerably tilted from
the mean π plane and the P-phenyl group leans outward from
the cavity. As a result of this large deviation, the C3-bridge
and P-Ph moieties in 6-S are located in regions where the
diatropic ring-current effect is weak. However, the highly
distorted 22π circuit of 6-S retains moderate aromaticity;
significant upfield shifts (Δδ = 0.84-0.99 ppm) are seen for
the pyrrole-β protons of 6-S as compared with those of the
corresponding calixphyrin 8N-S24b,c (Table 1). Indeed, the
NICS value at the center of the two facing nitrogen atoms
of 6-S-m (-8.8) is only slightly positive relative to that
of 6-O-m (-9.9) (Table 4).
FIGURE 6. 1H NMR monitoring experiments of the reactions
between P,X,N2-porphyrins (3.3 μmol) and H2O2 (30 wt %, 3.6
μmol) in CDCl3/THF-d8 (total 0.64 mL; v/v = 14/1) with 1,1,2,2-
tetrachloroethane as an internal standard. (a) The reaction of 1N0
(open red circle); 1N0-O (pink circle), 120 (brown circle). (b) The
reaction of 1N (open red circle); 1N-O (pink circle), 12 (brown
circle), 6-O (purple circle). (c) The reaction of 1S (open red circle);
1S-O (pink circle), 7-O (green circle). Asterisk (/) indicates the
peaks of residual solvents.
In contrast to 6-E (E = O, S), 7-O displays weakly
1
paratropic ring current effects in the H NMR spectrum
(Figure 1e). Specifically, the pyrrole-β and P-phenyl protons
of 7-Oappear at δ4.99-5.68 and 7.75-8.64 ppm, respectively,
(44) The phosphorus atom deviates from the respective mean π planes by
To obtain more information on the structures and the
aromaticity of 6-O, 6-S, and 7-O, DFT calculations on
their model compounds 6-O-m, 6-S-m, and 7-O-m
(Chart 4) were performed at the level of B3LYP/6-311G(d,
p). The optimized structures of these models are depicted in
˚
˚
˚
˚
0.30 A (for 1N-m), 1.16 A (for 6-O-m), 1.35 A (for 6-S-m), and 1.24 A (for
7-O-m), and the P-C(Ph) bond axis slants against the plane by 84.7ꢀ (for
1N-m), 116.4ꢀ (for 6-O-m), 135.2ꢀ (for 6-S-m), and 113.9ꢀ (for 7-O-m).
(45) The N S distances for the NH S hydrogen bonds in silver-anti-
3 3 3
3 3 3
˚
mony sulfides were reported to be 3.20-3.65 A: Vaqueiro, P.; Chippindale,
A. M.; Cowley, A. R.; Powell, A. V. Inorg. Chem. 2003, 42, 7846–7851.
386 J. Org. Chem. Vol. 75, No. 2, 2010

Nakabuchi et al.
JOCArticle
SCHEME 4. DDQ Oxidation of P-Functionalized Porphyrinogens 5N-O, 5N-S, 5S-O, and 5S-S
SCHEME 5. Reconstruction of π Systems via P-Oxidation
which are slightly upfield shifted (Δδ= -0.35 to -0.97 ppm) or
downfield shifted (Δδ = 0.41-0.69 ppm) when compared with
the chemical shifts of the protons in the corresponding calix-
phyrin 1424b,c (Chart 3 and Table 1). Small, positive NICS
values (þ2.4) at the center of 7-O-m support that 7-O
has weak antiaromaticity in terms of magnetic criterion
(Table 4).
In the UV-vis absorption spectra of 6-E (E = O, S), split
Soret-like bands (6-O, λab 422 and 494 nm; 6-S, λab 424 and
485 nm) and a remarkably broad, low-energy band reaching
into a near-infrared region were observed (Figure 9a and
Table 3). The absorption coefficients of 6-S are consider-
ably smaller than that of 6-O. This is presumably due to the
structural deviation of 6-S. The UV-vis absorption
spectrum of 7-O shows a broad Soret-like band at the
high-energy region (λab 394 nm) and no detectable Q bands,
which is characteristic of highly ruffled, nonaromatic 4nπ
porphyrinoids (Figure 9b).25,46
The electrochemical oxidation processes of 6-E (E =
O, S) were found to be irreversible, whereas the reduction
processes were reversible or quasi-reversible (Figure S2
in Supporting Information). The first oxidation and reduc-
tion potentials of 6-O (Eox,1 = þ0.10 V; Ered,1 = -1.21 V,
vs Fc/Fcþ) and 6-S (Eox,1 = þ0.21 V; Ered,1 = -1.13 V, vs
Fc/Fcþ) are more cathodic and more anodic as compared
FIGURE 7. Top (upper) and side (lower) views of 7-O (30%
probability ellipsoids). Hydrogen atoms (except for NH) and
solvent molecules are omitted for clarity.
with the respective values of 1N (Figure 3). As a result,
this yields very small HOMO-LUMO gaps for 6-O (ΔE =
1.31 V) and 6-S (ΔE = 1.34 V) in comparison with 1N
(ΔE = 1.89 V). The electrochemical oxidation processes of
7-O occurred reversibly at -0.29 and þ0.02 V (vs Fc/Fcþ).
The first oxidation and reduction potentials of 7-O
(Eox,1 = -0.29 V; Ered,1 = -1.92 V, vs Fc/Fcþ) are consider-
ably cathodically shifted when compared with the respective
values of 1S (Figure 3), which exhibits the isophlorin character
of the π system of 7-O. As a whole, the calculated HOMO and
LUMO levels of the model compounds 6-O-m (-4.96 and
-3.03 eV), 6-S-m (-5.00 and -3.12 eV), and 7-O-m (-4.47
and -2.41 eV) are in good agreement with the observed redox
potentials (Figure S3 in Supporting Information). It is note-
worthy that the HOMO-LUMO gaps of the P-oxidized
derivatives 6-O, 6-S, and 7-O are much narrower than those
(46) (a) Yamamoto, Y.; Yamamoto, A.; Furuta, S.-y.; Horie, M.;
Kodama, M.; Sato, W.; Akiba, K.-y.; Tsuzuki, S.; Uchimaru, T.; Hashizume,
D.; Iwasaki, F. J. Am. Chem. Soc. 2005, 127, 14540–14541. (b) Cissell, J. A.;
Vaid, T. P.; Yap, G. P. A. Org. Lett. 2006, 8, 2401–2404. (c) Liu, C.; Shen,
D.-M.; Chen, Q.-Y. J. Am. Chem. Soc. 2007, 129, 5814–5815.
J. Org. Chem. Vol. 75, No. 2, 2010 387

Nakabuchi et al.
JOCArticle
FIGURE 8. Top (upper) and side (lower) views of the optimized structures for (a) 6-O-m, (b) 6-S-m, and (c) 7-O-m: gray (C), light blue (H),
blue (N), red (O), orange (P), yellow (S).
the structures of the resulting P-oxygenated porphyrins strongly
depend on the combination of the core heteroatoms and the
structure of the peripherally fused carbocycles. The present
results demonstrate for the first time that the introduction of a
phosphorus atom into the core is a highly promising methodol-
ogy to modify the fundamental properties and reactivities of
porphyrin π systems.
Experimental Section
X-ray Crystallography. Single crystals suitable for X-ray
analysis were grown from CH2Cl2-MeOH (for 1N, 6-O, and
S5) or from CH2Cl2 (for 7-O). X-ray crystallographic measure-
ments were made on a Rigaku Saturn CCD area detector
˚
with graphite monochromated Mo KR radiation (0.71070 A)
at -150 ꢀC. The data were corrected for Lorentz and polariza-
tion effects. The structures were solved by using direct meth-
ods47 and refined by full-matrix least-squares techniques against
F2 using SHELXL-97.48 The non-hydrogen atoms were refined
anisotropically, and hydrogen atoms were refined using the rigid
model. The CIF files of 1N and 6-O were deposited as
Supporting Information files in ref 23b (1N and 6-O corre-
spond to 7 and 8, respectively, in ref 23b).
DFT Calculations on Model Compounds. The structures of
FIGURE 9. UV-vis absorption spectra of (a) 6-O (reddish-
model compounds were optimized using DFT. The basis
set used was 6-311G(d,p).49 The functionals of DFT was the
Becke 1988 exchange and Lee-Yang-Parr correlation func-
tionals (B3LYP).50 The NICS values51 were calculated at the
Hartree-Fock level with gauge-including atomic orbitals
(GIAOs) at the DFT optimized structures. The basis set
used in the NICS value computations was 6-31þG(d).52 All
purple) and 6-S (bluish-purple) and (b) 7-O (green) in CH2Cl2.
The spectra of corresponding 18π porphyrins 1N and 1S are shown
by a black line in each spectrum.
of the 18π σ3-P,X,N2 porphyrins 1X. Most importantly, the
optical and electrochemical properties of the P,X,N2-porphyrin
systems are readily tunable by simple P-oxidation reactions.
5. Concluding Remarks. We have successfully prepared 18π
P,X,N2-porphyrins 1N,1N0,and1S via an acid-promoted [3 þ 1]
condensation approach using phosphatripyrranes and 2,5-bis-
[hydroxy(phenyl)methyl]heteroles. Both experimental (NMR,
UV-vis, and X-ray) and theoretical (DFT calculations) results
have revealed that these P,X,N2-porphyrins possess reasonably
high aromaticity that stem from the slightly deviated
18π-annulene circuit. The UV-vis and CV/DPV measurements
as well as the DFT calculations have also revealed that these P,X,
N2-porphyrins possess considerably narrow HOMO-LUMO
gaps. It is of particular interest that the simple P-oxygenation of
σ3-P,X,N2-porphyrins with H2O2 causes unprecedented types of
π-reconstruction from 18π into peripherally extended 22π or
isophlorin-type 20π circuits. Notably, the reaction modes and
(47) SIR92: Altomare, A.; Cascarano, G.; Giacovazzo, C.; Guagliardi,
A.; Burla, M.; Polidori, G.; Camalli, M. J. Appl. Crystallogr. 1994, 27,
435-436.
€
(48) Sheldrick, G. M. SHELXL-97; University of Gottingen: Germany,
1997.
(49) (a) Krishnan, R.; Binkley, J. S.; Seeger, R.; Pople, J. A. J. Chem.
Phys. 1980, 72, 650–654. (b) McLean, A. D.; Chandler, G. S. J. Chem. Phys.
1980, 72, 5639–5648.
(50) (a) Becke, A. D. J. Chem. Phys. 1988, 98, 5648–5652. (b) Lee, C.;
Yang, W.; Parr, R. G. Phys. Rev. B 1988, 37, 785–789.
(51) Schleyer, P. v. R.; Maerker, C.; Dransfeld, A.; Jiao, H.; van Eikema
Hommes, N. J. R. J. Am. Chem. Soc. 1996, 118, 6317–6318.
(52) (a) Hehre, W. J.; Ditchfield, R.; Pople, J. A. J. Chem. Phys. 1972, 56,
2257–2261. (b) Francl, M. M.; Pietro, W. J.; Hehre, W. J.; Binkley, J. S.;
DeFrees, D. J.; Pople, J. A.; Gordon, M. S. J. Chem. Phys. 1982, 77, 3654–
3665. (c) Clark, T.; Chandrasekhar, J.; Spitznagel, G. W.; Schleyer, P. von R.
J. Comput. Chem. 1983, 4, 294–301.
(53) Pople, J. A. et al. Gaussian 03; Gaussian, Inc.: Pittsburgh PA, 2003.
For a full list of the authors, see Supporting Information.
388 J. Org. Chem. Vol. 75, No. 2, 2010

Nakabuchi et al.
JOCArticle
calculations were carried out using the Gaussian 03 suite of
programs.53 In the optimized structures of model compounds
except for 1S-m, the P-phenyl group is parallel to the P-X axis
(X = NH, S). In the optimized structure of 1S-m, the P-phenyl
ring is found to be vertical to the P-S axis.23a However, the
difference in energy between the optimized structure and the
structure in which the P-phenyl ring is rotated by 90ꢀ (i.e., the P-
phenyl ring is parallel to the P-S axis) is very small (1.87 kcal
mol-1). Thus, the binding for the P-phenyl ring rotation is
considered to be very weak. In this paper, the parallel structure
was used for discussion.
the Ministry of Education, Culture, Sports, Science and Tech-
nology of Japan. We thank Prof. Takahiro Sasamori, Prof.
Norihiro Tokitoh (Kyoto University), and Prof. Hidemitsu Uno
(Ehime University) for the analysis of X-ray crystallography. We
also thank Dr. Kengo Suzuki (Hamamatsu Photonics K.K.) for
the measurement of absolute fluorescence quantum yields of 2S.
T.N. thanks a JSPS fellowship for young scientists.
Supporting Information Available: Experimental details and
1
characterization data, H and 13C NMR charts for new com-
pounds, CIF files for 7-O and S5, Cartesian coordinates and
energies of B3LYP/6-311G(d,p) geometry optimized structures
for model compounds, and complete ref 53. This material is
available free of charge via the Internet at http://pubs.acs.org.
Acknowledgment. This work was supported by a Grant-in-
Aid for Science Research on Priority Areas (no. 20038039) from
J. Org. Chem. Vol. 75, No. 2, 2010 389