Three-coordinate, phosphine-ligated azadipyrromethene complexes of univalent group 11 metals

Article abstract of DOI:10.1021/ic900208a

Tetraarylazadipyrromethenes are Lewis basic, red-light absorbing dyes with optical properties conducive to sensing and therapeutic applications. Recently, transition metal complexes of these ligands have been described. Here, we report a series of three-c

Full text of DOI:10.1021/ic900208a

8134 Inorg. Chem. 2009, 48, 8134–8144
DOI: 10.1021/ic900208a
Three-Coordinate, Phosphine-Ligated Azadipyrromethene Complexes of Univalent
Group 11 Metals
Thomas S. Teets,† James B. Updegraff III,† Arthur J. Esswein,‡ and Thomas G. Gray*,†
†Department of Chemistry, Case Western Reserve University, Cleveland, Ohio 44106, and ‡Department of
Chemistry, Massachusetts Institute of Technology, Cambridge, Massachusetts 02139
Received February 2, 2009
Tetraarylazadipyrromethenes are Lewis basic, red-light absorbing dyes with optical properties conducive to sensing and
therapeutic applications. Recently, transition metal complexes of these ligands have been described. Here, we report a series
of three-coordinate Group 11 complexes of unsubstituted and methoxy-substituted tetraarylazadipyrromethenes. In each, two
pyrrole nitrogens chelate a d10 metal ion; triphenyl- or triethylphosphine occupies a third coordination site. New complexes are
characterized by multinuclear NMR, X-ray crystallography, optical absorption and emission spectroscopy, and elemental
analysis. Solid-state structures show trigonal planar geometries about the metal centers, and reveal pervasive intra- and
intermolecular π-stacking interactions. Visible light absorption intensifies with metal binding, in some cases shifting to longer
wavelengths. The complexes weakly luminesce in the red region; emission wavelengths and quantum yields are similar to
those of free azadipyrromethenes. Methoxy-substitution on the ligand red-shifts optical features, whereas substitution of
triethylphosphine for triphenylphosphine in the third coordination site has minimal structural or spectral consequences.
Introduction
because of the chromophores’ wide optical tunability. Remote
functionalization through a backbone phenyl ring (at C-5,
Scheme 1a) allows chemical attachment with minimal changes
in optical properties. Reports of nucleotide, protein, and lipid
tagging with these versatile fluorophores are now common.5-20
Dipyrromethenes and azadipyrromethenes are chromo-
phoric, bidentate Lewis bases of great importance because of
their boron adducts. Scheme 1 depicts the general structures
þ
of dipyrromethene and azadipyrromethene BF2 chelates.
Boron dipyrromethenes and azadipyrromethenes absorb
visible light and yield intense fluorescence; their optical
attributes prompt continuing scrutiny.
(5) Chang, T.-C.; Lin, J.-J.; Lin, K.-C.; Lin, Y.-C.; Huang, W.-C.; Yang,
Y.-P.; Cheng, J.-Y. J. Lumin. 2002, 98, 149–152.
(6) Gite, S.; Mamaev, S.; Olejnik, J.; Rothschild, K. Anal. Biochem. 2000,
279, 218–225.
Difluoroboron dipyrromethenes are marketed as BODIPY
dyes.1 These absorb light at a range of frequencies (ca. 500-
650 nm in methanol) depending on peripheral substitution.
Stokes shifts are small, and an emission spectrum that mirrors
the principal absorption band reaches a maximum at wave-
lengths 10-20 nm red-shifted from the absorption peak.
Emission persists with 4 ns or longer lifetimes, and fluorescence
quantum yields often exceed 80%. The BODIPY chromo-
phore is non-ionizable, and optical characteristics are mini-
mally sensitive to pH or solvent polarity. Substitution is
possible through the pyrrolic carbon atoms, at the backbone
methine carbon, or along phenyl substituents at the methine
carbon. Much current research explores substitution at boron,
in some cases with moieties that themselves fluoresce.2-4
Applications to bioconjugate chemistry are legion, not least
(7) Epand, R. F.; Sayer, B. G.; Epand, R. FEBS J. 2005, 272, 1792–1803.
(8) Teske, C. A.; Schroeder, M.; Simon, R.; Hubbuch, J. J. J. Phys. Chem.
B 2005, 109, 13811–13817.
(9) Fernandes, F.; Loura, L. M. S.; Prieto, M.; Koehorst, R.; Spruijt, R.
B.; Hemminga, M. A. Biophys. J. 2003, 85, 2430–2441.
(10) Sculimbrene, B.; Imperiali, B. J. Am. Chem. Soc. 2006, 128, 7346–
7352.
(11) Coleman, M. A.; Lao, V. H.; Segelke, B. W.; Beernink, P. T.
J. Proteome Res. 2004, 3, 1024–1032.
::
(12) Janosch, S.; Nicolini, C.; Ludolph, B.; Peters, C.; Volkert, M.;
Hazlet, T. L.; Waldmann, H.; Winter, R. J. Am. Chem. Soc. 2004, 126,
7496–7503.
::
::
(13) Kruger, P.; Baatz, J. E.; Dluhy, R. A.; Losche, M. Biophys. Chem.
2002, 99, 209–228.
(14) Reaven, E.; Leers-Sucheta, S.; Nomoto, A.; Azhar, S. Proc. Natl.
Acad. Sci. U.S.A. 2001, 98, 1613–1618.
::
:: ::
˚
(15) Bergstrom, F.; Hagglof, P.; Karolin, J.; Ny, T.; Johansson, L. B.-A.
Proc. Natl. Acad. Sci. U.S.A. 1999, 96, 12477–12481.
(16) Yee, M.-c.; Fas, S. C.; Stohlmeyer, M. M.; Wandless, T. J.; Cimprich,
K. A. J. Biol. Chem. 2005, 280, 29053–29059.
(17) Karolin, J.; Johansson, L. B.-A.; Strandberg, L.; Ny, T. J. Am. Chem.
Soc. 1994, 116, 7801–7806.
(18) Tarahovsky, Y. S.; Koynova, R.; MacDonald, R. C. Biophys. J.
2004, 87, 1054–1064.
*To whom correspondence should be addressed. E-mail: tgray@case.edu.
(1) BODIPY is a registered trademark of Molecular Probes, Inc.
(2) Ulrich, G.; Goze, C.; Guardigli, M.; Roda, A.; Ziessel, R. Angew.
Chem., Int. Ed. 2005, 44, 3694–3698.
(3) Goze, C.; Ulrich, G.; Mallon, L. J.; Allen, B. D.; Harriman, A.;
Ziessel, R. J. Am. Chem. Soc. 2006, 128, 10231–10239.
(19) Epand, R.; Vuong, P.; Yip, C. M.; Maekawa, S.; Epand, R. F.
Biochem. J. 2004, 379, 527–532.
(20) Malinin, V. S.; Haque, Md. E.; Lentz, B. R. Biochemistry 2001, 40,
8292–8299.
ꢀ
(4) Ziessel, R.; Goze, C.; Ulrich, G.; Cesario, M.; Retailleau, P.;
Harriman, A.; Rostron, J. P. Chem.;Eur. J. 2005, 11, 7366–7378.
r
pubs.acs.org/IC
Published on Web 08/05/2009
2009 American Chemical Society

Article
Inorganic Chemistry, Vol. 48, No. 17, 2009 8135
Scheme 1
quantum yields can exceed 40%.43 Introducing bromine
atoms along the azadipyrromethene periphery begets tri-
plet-state photochemistry. Photosensitization of O2 occurs,
and in vitro photokilling of human cervical carcinoma
(HeLa) cells was reported.44,45 These reports, combined with
the red-absorption of boron azadipyrromethenes, suggest
azadipyrromethenes as potent mediators of photodynamic
therapy for cancers and other afflictions.
In the synthesis of boron dipyrromethenes and azadipyr-
romethenes, the conjugated skeleton is prepared first, and
boron is added last. Clear opportunities exist for metal-ion
binding, where the chromophore acts as a bidentate chelate.
Metallocomplexes of dipyrromethenes have been investi-
gated at some length; an emphasis has been supramolecular
chemistry. Cohen and co-workers report six-coordinate tris-
(dipyrromethene) complexes of first-row transition ele-
ments46-52 and group 13 ions.53 In many such compounds
a backbone meso-phenyl group bears Lewis basic sites.
Porous coordination polymers form upon reaction with
soluble metal-ion sources. Ultraviolet excitation of tris-
(5-mesityldipyrromethene) complexes of gallium and indium
elicits green fluorescence with ns-scale lifetimes. Dolphin and
collaborators54,55 have published accounts of R- and β-linked
dipyrromethenes in the construction of metallohelices and
cyclic trimers of metal complexes. Bis(dipyrromethene)zinc-
(II) species have luminescence properties that are sensitive to
rotational freedom of the meso substituent.56 For example,
bis(5-mesityldipyrromethene)zinc(II) emits with a fluore-
scence quantum yield Φem = 0.36 at room temperature in
toluene, whereas the 5-phenyl analogue is minimally fluore-
scent, Φem = 0.006.57 Bis(dipyrromethene)zinc complexes
act as energy-transfer donors to zinc porphyrins in light-
harvesting arrays.58 Maeda and co-workers59 have disclosed
the nanoscale morphologies of (dipyrrinato)zinc complexes
bearing meso-phenylene ethynylene moieties; some are lumi-
nescent.
BODIPY dyes serve as energy-transfer donors to porphy-
rins21-23 and other acceptors24-28 in light-harvesting arrays.
Ultrafast electron-transfer studies across supramolecular as-
semblies have appeared where BODIPY moieties act as elec-
tron donors29 and acceptors.30,31 BODIPY moieties having
pendant Lewis bases, including amines, thioethers, and thia-
crowns, are fluorescence metal-ion and NO sensors.32-41
A
recent account42 describes triplet-state photosensitization with
a di-iodo boron dipyrromethene, brought about by the heavy-
atom effect of iodine, which enhances excited-state intersystem
crossing.
Boron azadipyrromethenes (Scheme 1b) are fast regaining
attention for their photophysical properties. Their signature
optical features are absorptions centered near 300 and
620 nm with molar absorptivities in the tens of thousands.
The longer-wavelength absorption falls just within an energy
range suitable for human photodynamic therapy. Botþh
transitions have interest for solar-energy capture. BF2
chelates of azadipyrromethenes are fluorescent, and emission
(21) Wagner, R. W.; Lindsey, J. S. Pure Appl. Chem. 1996, 68, 1373–1380.
(22) Wagner, R. W.; Lindsey, J. S.; Seth, J.; Palaniappan, V.; Bocian, D.
F. J. Am. Chem. Soc. 1996, 118, 3996–3997.
(23) Li, F.; Yang, S. I.; Ciringh, Y.; Seth, J.; Martin, C. H.; Singh, D. L.;
Kim, D.; Birge, R. R.; Bocian, D. F.; Holten, D.; Lindsey, J. S. J. Am. Chem.
Soc. 1998, 120, 10001–10017.
(24) Yilmaz, M. D.; Bozdemir, O. A.; Akkaya, E. U. Org. Lett. 2006, 8,
2871–2873.
(25) Hepp, A.; Ulrich, G.; Schmechel, R.; von Seggern, H.; Ziessel, R.
Synth. Met. 2004, 146, 11–15.
More recently, reports of metal complexes of azadipyrro-
methenes have emerged. We have described homoleptic,
(26) Herrmann, A.; Weil, T.; Sinigersky, V.; Wiesler, U.-M.; Vosch, T.;
Hofkens, J.; De Schryver, F. C.; Mullen, K. Chem.;Eur. J. 2001, 7, 4844–
4853.
ꢀ
(27) Serin, J. M.; Brousmiche, D. W.; Frechet, J. M. J. Chem. Commun.
(43) Loudet, A.; Bandichhor, R.; Wu, L.; Burgess, K. Tetrahedron 2008,
64, 3642–3654.
(44) Killoran, J.; Allen, L.; Gallagher, J. F.; Gallagher, W. M.; O’Shea, D.
F. Chem. Commun. 2002, 1862–1863.
(45) Gorman, A.; Killoran, J.; O’Shea, C.; Kenna, T.; Gallagher, W. M.;
O’Shea, D. F. J. Am. Chem. Soc. 2004, 126, 10619–10631.
(46) Halper, S. R.; Do, L.; Stork, J. R.; Cohen, S. M. J. Am. Chem. Soc.
2006, 128, 15255–15268.
(47) Halper, S. R.; Cohen, S. M. Inorg. Chem. 2005, 44, 486–488.
(48) Murphy, D. L.; Malachowski, M. R.; Campana, C. F.; Cohen, S. M.
Chem. Commun. 2005, 5506–5508.
2002, 2605–2607.
ꢀ
(28) Serin, J. M.; Brousmiche, D. W.; Frechet, J. M. J. J. Am. Chem. Soc.
2002, 124, 11848–11849.
(29) Harriman, A.; Rostron, J. P.; Cesario, M.; Ulrich, G.; Ziessel, R.
J. Phys. Chem. A 2006, 110, 7994–8002.
(30) Hattori, S.; Ohkubo, K.; Urano, Y.; Sunahara, H.; Nagano, T.;
Wada, Y.; Tkachenko, N. V.; Lemmetyinen, H.; Fukuzumi, S. J. Phys.
Chem. B 2005, 109, 15368–15375.
(31) Kollmannsberger, M.; Rurack, K.; Resch-Genger, U.; Daub, J.
J. Phys. Chem. A 1998, 102, 10211–10220.
(32) Zeng, L.; Miller, E. W.; Pralle, A.; Isacoff, E. Y.; Chang, C. J. J. Am.
Chem. Soc. 2006, 128, 10–11.
(33) Ueno, T.; Urano, Y.; Kojima, H.; Nagano, T. J. Am. Chem. Soc.
2006, 128, 10640–10641.
(34) Gabe, Y.; Ueno, T.; Urano, Y.; Kojima, H.; Nagano, T. Anal.
Bioanal. Chem. 2006, 386, 621–626.
(35) Gabe, Y.; Urano, Y.; Kikuchi, K.; Kojima, H.; Nagano, T. J. Am.
Chem. Soc. 2004, 126, 3357–3367.
(36) Coskun, A.; Akkaya, E. U. J. Am. Chem. Soc. 2005, 127, 10464–
10465.
(37) Turfan, B.; Akkaya, E. U. Org. Lett. 2002, 4, 2857–2859.
(38) Ulrich, G.; Ziessel, R. J. Org. Chem. 2004, 69, 2070–2083.
(39) Domaille, D. W.; Que, E. L.; Chang, C. J. Nat. Chem. Biol. 2008, 4,
168–175.
(49) Halper, S. R.; Cohen, S. M. Inorg. Chem. 2005, 44, 4139–4141.
(50) Halper, S. R.; Malachowski, M. R.; Delaney, H. M.; Cohen, S. M.
Inorg. Chem. 2004, 43, 1242–1249.
(51) Halper, S. R.; Cohen, S. M. Chem.;Eur. J. 2003, 9, 4661–4669.
(52) Cohen, S. M.; Halper, S. R. Inorg. Chim. Acta 2002, 341, 12–16.
(53) Thoi, V. S.; Stork, J. R.; Magde, D.; Cohen, S. M. Inorg. Chem. 2006,
45, 10688–10697.
(54) Zhang, Y.; Thompson, A.; Rettig, S. J.; Dolphin, D. J. Am. Chem.
Soc. 1998, 120, 13537–13538.
(55) Thompson, A.; Rettig, S. J.; Dolphin, D. Chem. Commun. 1999, 631–
632.
(56) Sutton, J. M.; Rogerson, E.; Wilson, C. J.; Sparke, A. E.; Archibald,
S. J.; Boyle, R. W. Chem. Commun. 2004, 1328–1329.
(57) Sazanovich, I. V.; Kirmaier, C.; Hindin, E.; Yu, L.; Bocian, D. F.;
Lindsey, J. S.; Holten, D. J. Am. Chem. Soc. 2004, 126, 2664–2665.
(58) Yu, L.; Muthukumaran, K.; Sazanovich, I. V.; Kirmaier, C.; Hindin,
E.; Diers, J. R.; Boyle, P. D.; Bocian, D. F.; Holten, D.; Lindsey, J. S. Inorg.
Chem. 2003, 42, 6629–6647.
(40) Rurack, K.; Kollmannsberger, M.; Resch-Genger, U.; Daub, J.
J. Am. Chem. Soc. 2000, 122, 968–969.
ꢀ
(41) Baruah, M.; Qin, W.; Vallee, R. A. L.; Beljonne, D.; Rohand, T.;
Dehaen, W.; Boens, N. Org. Lett. 2005, 7, 4377–4380.
(42) Yogo, T.; Urano, Y.; Ishitsuka, Y.; Maniwa, F.; Nagano, T. J. Am.
Chem. Soc. 2005, 127, 12162–12163.
(59) Maeda, H.; Hasegawa, M.; Hashimoto, T.; Kakimoto, T.; Nishio, S.;
Nakanisi, T. J. Am. Chem. Soc. 2006, 128, 10024–10025.

8136 Inorganic Chemistry, Vol. 48, No. 17, 2009
Teets et al.
bis-chelate complexes of ZnII and HgII.60 The zinc complexes
share several structural commonalities with dipyrromethene
complexes, though the dihedral angle of the two chelate rings
is strongly influenced by intramolecular π-π interactions
between a pyrrole and a substituent phenyl ring. The four-
coordinate group 12 complexes display more intense visible
absorption maxima that are red-shifted relative to the free
ligands; the complexes are non-emissive at room temp-
erature. Complexes of the tetraphenylazadipyrromethene
with first-row ions CoII, NiII, and CuII, along with the ZnII
complex we reported, appeared later.61 As with ZnII, com-
plexation of these other divalent first-row metals results in
pronounced changes in the absorption spectra.
A second class of azadipyrromethene coordination com-
pounds are the three-coordinate, group 11 complexes dis-
closed earlier.62 These contain a d10 MI (M = Cu, Ag, Au)
center chelated by tetraphenylazadipyrromethene, with tri-
phenylphosphine (PPh3) bound to the third coordination
site, resulting in an essentially trigonal planar geometry. The
low-energy absorption maximum intensifies and red-shifts in
the group 11 complexes relative to the parent ligand. All three
are weakly emissive at room temperature, displaying emis-
sion maxima of ∼650 nm and <1% quantum yields. Pre-
liminary density functional theory (DFT) calculations on the
silver complex indicate primarily ligand-centered frontier
orbitals and electronic transitions.
Here, we detail the synthesis, characterization, and photo-
physical properties of an extended series of group 11 azadi-
pyrromethene complexes. All but one are structurally
characterized by single crystal X-ray diffraction, whereas
spectral methods and microanalysis verify the purity of all
products. Included are CuI and AgI complexes with methoxy-
substituted azadipyrromethenes; substitution of the pendant
aryl rings perturbs the complexes’ absorption spectra. We
also describe AgI derivatives where triethylphosphine re-
places triphenylphosphine as the capping ligand. We enu-
merate the structural and spectral consequences of inserting
this smaller phosphine, which cannot participate in the
intramolecular π-π interactions that are common in PPh3-
bound complexes. Room temperature absorption and emis-
sion spectra are reported for all compounds. The trends in the
absorption and emission spectra noted earlier,60 namely, the
intensified and red-shifted maxima relative to free ligand,
recur in the derivatives described here. DFT and time-
dependent DFT calculations find the visible absorptions of
azadipyrromethene complexes to be ligand-centered; metal-
ion involvement is peripheral.
Scheme 2
azadipyrromethene in the presence of diisopropylethyla-
mine. (Triphenylphosphine)copper(I) triflate was gener-
ated in situ from free triphenylphosphine and the toluene
adduct of copper(I) triflate. Likewise, (triethylphos-
phine)silver(I) triflate was prepared in situ from silver
triflate and a slight excess of triethylphosphine. As de-
scribed previously,62 AuI complex 9 is prepared by in situ
deprotonation of tetraphenylazadipyrromethene, fol-
lowed by metathesis with AuI(PPh3)Cl. Tetrahydrofuran
(THF) solutions of the free azadipyrromethene are dark
blue; upon metalation, the color changes to a somewhat
brighter blue. Syntheses of CuI and AgI complexes are
complete in 60 min, as judged by NMR, and isolated
yields range from 62-89%. All complexes have been
characterized by 1H and 31P NMR spectroscopy, absorp-
tion and emission spectroscopy, and elemental analysis;
crystal structures of eight have been determined.
Azadipyrromethene 1H NMR spectra are largely insen-
sitive to copper(I) or silver(I) binding.62 The lone excep-
tions are the resonances associated with the azadipyrro-
methene phenyl rings closest to the metal-phosphine
fragment in triphenylphosphine complexes. These reso-
nances shift upfield by 0.3-0.7 ppm upon complexation,
and the methoxy singlet in bound Lb also shifts upfield by
a similar amount for 2 and 5. Resonances of triphenyl-
phosphine ligands in 1-6 and 9 are essentially invariant;
they overlap with azadipyrromethene resonances. The
31P{1H} NMR spectra show significantly broadened
resonances, and in the silver complexes the expected
doublet of doublets splitting pattern, arising from 31P
coupling to 107Ag and 109Ag (51.8% and 48.2% abun-
dant, respectively), is unresolved.
Crystal Structures. Structures of 1-5 and 7-9 have
been determined crystallographically. Those of 1, 4, and 9
were reported previously.62 No complex possesses crys-
tallographically imposed symmetry. All structures con-
tain discrete molecules with three-coordinate metal
centers. Coordination geometries are approximately tri-
gonal planar, though in all cases the N;M;N bond
angles are decidedly smaller than the N;M;P angles.
Thermal ellipsoid projections for PPh3CuLb (2),
PPh3AgLb (5), and PEt3AgLc (8) appear in Figure 1.
Table 2 summarizes crystallographic data for the new
azadipyrromethene complexes reported here.
Results and Discussion
Syntheses. Scheme 2 details the synthesis of the azadi-
pyrromethene complexes considered here; Table 1 col-
lects optimized yields. Reaction conditions were analo-
gous to those previously employed in the syntheses of
PPh3CuLa (1) and PPh3AgLa (4).62 Complexes of copper-
(I) and silver(I) assembled rapidly in room-temperature
reactions of (phosphine)metal(I) triflates with the free
Table 3 collects selected average interatomic distances
and angles for the complexes of CuI(1-3) and AgI(4, 5, 7,
and 8), as well as the corresponding distances for
AuI complex 9. Metal-nitrogen bond distances range
(60) Teets, T. S.; Partyka, D. V.; Updegraff, J. B., III; Gray, T. G. Inorg.
Chem. 2008, 47, 2338–2346.
::
I
(61) Palma, A.; Gallagher, J. F.; Muller-Bunz, H.; Wolowska, J.;
˚
from 1.9784(13)-1.9914(14) A (M = Cu ), 2.2073(12)-
McInnes, E. J. L.; O’Shea, D. F. Dalton Trans. 2009, 273–279.
(62) Teets, T. S.; Partyka, D. V.; Esswein, A. J.; Updegraff, J. B., III;
Zeller, M.; Hunter, A. D.; Gray, T. G. Inorg. Chem. 2007, 46, 6218–6220.
˚
2.2811(19) A (M = Ag), and 2.2349(12) and 2.2448(12) A
˚
(M = Au, 9). In all compounds, slight asymmetries in

Article
Inorganic Chemistry, Vol. 48, No. 17, 2009 8137
Figure 1. X-ray crystal structures of 2, 5, and 8. Thermal ellipsoids appear at the 50% probability level; data were collected at 100 K. Hydrogen atoms are
omitted for clarity, and carbon atoms, shown in gray, are unlabeled.
Table 1. Syntheses and Designation of Compounds
M-N bond lengths occur. The most asymmetric
(azadipyrromethene)MI grouping, in terms of bond
lengths to nitrogen, is one crystallographically indepen-
dent molecule of 5, where the discrepancy in Ag-N bond
here are not unusual. For example, copper(I)-nitrogen
bond lengths in bis(2,9-diphenyl-3,4,7,8-tetramethyl-
1,10-phenanthroline)copper(I) tetraphenylborate range
63
˚
from 1.993(3)-2.150(3) A and those in a bis(2,9-pen-
tafluorophenyl-1,10-phenanthroline)copper(I) cation fall
˚
lengths is 0.07 A. Compound 3 has the most nearly equal
M-N bond lengths with the difference being 0.003 A.
between 2.048(3) and 2.109(3) A.64 In mixed phenanthro-
˚
˚
Metal-phosphorus bond lengths in homologous com-
pounds are 2.1800(5) A (1), 2.3503(6) A (4, averaged over
line-chelating diphosphine copper(I) complexes, Cu-N
65
˚
˚
˚
bond lengths range from 2.064(3)-2.109(2) A. Struc-
tures of two three-coordinate (bipyridine)silver(I) phos-
˚
three independent molecules), and 2.2088(3) A (9). Cor-
responding bond lengths to phosphorus in other com-
pounds are similar; substitution of PPh3 in 4 with PEt3 in
7 does not significantly alter the Ag-P bond length.
Methoxy-substitution on the azadipyrromethene elicits
small and non-systematic changes in metal-ligand bond
lengths and angles. Structural comparisons with related
compounds suggest that the bond metrics encountered
phine cations are available. Silver-nitrogen bond lengths
(63) Cunningham, C. T.; Moore, J. T.; Cunningham, K. L. H.; Fanwick,
P. E.; McMillin, D. R. Inorg. Chem. 2000, 39, 3638–3644.
(64) Hamilton, C. W.; Laitar, D. S.; Sadighi, J. P. Chem. Commun. 2004,
1628–1629.
(65) Kuang, S.-M.; Cuttell, D. G.; McMillin, D. R.; Fanwick, P. E.;
Walton, R. A. Inorg. Chem. 2002, 41, 3313–3322.

8138 Inorganic Chemistry, Vol. 48, No. 17, 2009
Teets et al.
Table 2. Crystallographic Data for Phosphine-ligated Azadipyrromethene Complexes
2
3
5
7
8
Formula
fw
C52H41CuN3O2P
834.42
C52H41CuN3O2P
834.42
C52H41AgN3O2P
877.2
C38H37AgN3P
674.56
C40H41AgN3O2P
734.61
crystal system
space group
triclinic
P1
monoclinic
P21/n
12.1711(8)
14.1525(9)
23.9305(16)
triclinic
P1
triclinic
P1
triclinic
P1
˚
a, A
15.0069(3)
17.9641(4)
19.0956(4)
82.9370(10)
73.2040(10)
67.2250(10)
4543.68(17)
4
1.334
100(2)
0.564
1904
0.36 ꢀ 0.26 ꢀ 0.20
1.23, 28.39
22509
18999
1175
1.030
0.0355
0.0948
13.8159(5)
17.4106(6)
19.3928(7)
75.636(2)
88.549(2)
85.457(2)
4504.7(32)
4
1.296
100(2)
0.525
1808
0.50 ꢀ 0.30 ꢀ 0.13
1.48, 34.02
131614
36074
1067
1.079
0.0373
0.1120
11.3616(3)
11.4783(3)
13.0739(3)
88.607(1)
73.817(1)
73.256(1)
1565.10(7)
2
1.415
100(2)
0.726
675
0.17 ꢀ 0.14 ꢀ 0.13
1.95, 33.21
32210
11642
420
1.102
0.0382
0.1114
11.4051(8)
11.7105(9)
15.7693(12)
110.336(3)
102.065(3)
89.987(4)
1925.2(2)
2
1.267
100(2)
0.600
760
0.48 ꢀ 0.30 ꢀ 0.25
1.83, 27.50
30783
8775
429
1.064
0.0570
0.1491
˚
b, A
˚
c, A
R, deg
β, deg
γ, deg
92.4870(10)
3
˚
cell volume, A
Z
4118.2(5)
4
1.346
100(2)
0.616
1736
0.33 ꢀ 0.19 ꢀ 0.18
1.67, 27.50
47402
9426
534
Dcalcd, Mg m-3
T, K
μ, mm-1
F(000)
cryst size, mm3
θmin, θmax, deg
no. of reflns collected
no. of indep reflns
no. of refined params
goodness-of-fit on F2a
final R indicesb [I > 2σ(I)] R1
wR2
1.027
0.0300
0.0738
R indices (all data)
R1
wR2
0.0445
0.1014
0.0386
0.0782
0.0482
0.1053
0.0470
0.1005
0.0597
0.1528
P
P
P
P
a GOF = [ w(Fo2 - Fc2)2/(n - p)]1/2; n = number of reflections, p = number of parameters refined. b R1 = ( Fo| -|Fc )/ |Fo|; wR2 = [ w-
] .
P
4 1/2
(Fo2 - Fc2)2/ wFo
˚
the N-M-N chelate. Figure 2 depicts this canting for
silver(I) complex 8, where the interplanar angle is 27.0ꢀ.
The largest such tilt encountered here is 29.8ꢀ for gold
complex 9; the smallest is 18.3ꢀ for one crystallographi-
cally independent molecule of 4. Substitution of PPh3
(cone angle = 145ꢀ),69 with less bulky PEt3 (cone angle =
132ꢀ) does not decrease the tilt of the coordination plane,
as judged from a comparison of the structures of 4 and 7.
At least for this one example, ligand tilting likely results
from a size mismatch between the MI ion and the binding
cavity. Furthermore, we note that these angles vary
considerably across the series, even for complexes with
the same metal ion. The structures of 2, 4, and 5 contain
two, three, and two molecules per asymmetric unit,
respectively. In these instances, crystallographically in-
dependent molecules of the same complex possess dis-
parate tilting angles, differing by more than 2.7ꢀ. These
observations indicate that the structure is quite flexible,
and that the crystallographically observed canting angle
is not rigidly imposed.
Ligand bite angles, defined as the N-M-N angle for
bound azadipyrromethenes, are 94.49(6)ꢀ for 1, 86.24(7),
82.96(7), and 82.27(7) for the three crystallographically
independent molecules of 4, and 81.54(4)ꢀ for 9. The
range for all compounds is 81.51(4)ꢀ for 9 to 95.12(5) for
3, and a decrease is observed traversing down the group 11
series from copper(I) to silver(I) to gold(I). For com-
parison, the bite angle in bis(2,9-diphenyl-3,4,7,8-tetra-
methyl-1,10-phenanthroline)copper(I) tetraphenylborate
is 81.23(11)ꢀ,63 and this angle is representative of those
in other cuprous phenanthroline complexes.64,65,70 In the
Table 3. Selected Interatomic Distances (A) and Angles (deg) for Group 11
Azadipyrromethene Complexes 1-9
CuIa
AgIb
AuIc
d(M-N)
1.9854(13)
2.1776(4)
94.06(5)
2.2423(17)
2.3403(9)
83.11(6)
2.2398(12)
2.2088(3)
81.51(4)
d(M-P)
—(N-M-N)
a Average for 1-3. b Average for 4, 5, 7, and 8. c Complex 9; M-N
bonds are averaged.
in two salts of [Ag(bpy)(PPh3)]þ range from 2.251(3)-
2.316(4) A. In a dinuclear complex67 where 1,4-bis-
66
˚
(diphenylphosphino)butane bridges two Ag(bpy)I ca-
˚
tions, Ag-N bond lengths are 2.227(3) and 2.334(3) A.
Metal-ion binding dilates the backside C-Nmeso-C
linkage away from the ideal 120ꢀ for sp2-hybridization.
For example, these angles about the meso-nitrogen in Lc
complexes 3 and 8 are 127.60(13)ꢀ and 128.3(2)ꢀ, respec-
tively. For comparison, O’Shea and co-workers report a
C-Nmeso-C angle of 119.7(3)ꢀ for the corresponding
BF2þ-chelate of Lc. These ligand distortions suggest
strain in group 11 azadipyrromethene complexes, but
also that the ligand is pliant enough to accommodate
large cations such as AgI (which is bigger than AuI).68
The metal-ion coordination geometry in 1-5 and 7-9
is trigonal planar. The coordination plane tilts relative to
the best-fit plane of the central azadipyrromethene ske-
leton. This tilting can be described as the angle between
the mean plane of the azadipyrromethene core (the two
pyrroles and bridging nitrogen) and the plane defined by
(66) Khalaji, A. D.; Amirnasr, M.; Aoki, K. Anal. Sci. 2006, 22, x87–x88.
(67) Zhang, L.; Chen, C.; Zhang, Q.; Zhang, H.; Kang, B. Acta Crystal-
logr. 2003, E59, m536–m537.
(68) Bayler, A.; Schier, A.; Bowmaker, G. A.; Schmidbaur, H. J. Am.
Chem. Soc. 1996, 118, 7006–7007.
(69) Tolman, C. A. Chem. Rev. 1977, 77, 313–348.
(70) Miller, M. T.; Gantzel, P. K.; Karpishin, T. B. Inorg. Chem. 1998, 37,
2285–2290.

Article
Inorganic Chemistry, Vol. 48, No. 17, 2009 8139
Figure 2. View of the crystal structure of 8 showing the canting of the
N-Ag-N plane relative to the mean plane of the ligand. The plane
depictedin redindicatesthe meanplane ofthe azadipyorromethene ligand
(the two pyrrole rings and the bridging nitrogen).
three-coordinate silver(I) complex [Ag(bpy)(PPh3)]2þ, the
N-Ag-N bite angle is 73.0(1)ꢀ for both the perchlorate
and hexafluorophosphate salts.66 In the disilver variant
[Ag2(bpy)2μ2-(dppb)](BF4)2, the corresponding bite angle
is 72.08(10)ꢀ.67 At parity of metal ion, the trend is toward
larger bite angles in azadipyrromethene ligands than for
either bipyridyl or phenanthroline in analogous com-
plexes. Metal-nitrogen bond lengths (vide supra) are
similar for diimine and azadipyrromethene ligands.
Figure 3. Room-temperature absorption spectra of La (red, dashed
line), 1 (light blue), 4 (green), and 9 (magenta), recorded in CHCl3.
Table 4. Summary of Absorption Wavelengths (nm) and Molar Absorptivities
(M-1 cm-1) for all Azadipyrromethene Ligands and Complexesa
λmax, ε (near-UV)
λmax, ε (red)
Inter- and intramolecular π-stacking is prominent in
metalla-azadipyrromethene crystal structures. Among
triphenylphosphine complexes, π-stacking of a phos-
phorus phenyl substituent and one proximal benzene ring
of the azadipyrromethene recurs. For example, the cen-
troid of one 4-methoxyphenyl group in an independent
molecule of 2 approaches the best-fit plane of a phosphine
La
Lb
Lc
1
2
3
4
5
6
7
303, 27 000
278, 68 000
291, 36 000
305, 38 000
320, 24 000
306, 32 000
306, 30 000
324, 30 000
304, 39 000
304, 30 000
303, 31 000
304, 40 000
597, 29 000
619, 37 000
608, 36 000
597, 38 000
620, 28 000
606, 38 000
615, 65 000
631, 69 000
615, 64 000
610, 42 000
614, 46 000
600, 45 000
˚
phenyl group (in the same molecule) within 3.53 A.
Similar stacking occurs in other structures. In the crystal
structure of 1, two distal phenyl groups in molecules
related by an inversion center approach with centroids
8
9
˚
a Measured in CHCl3 at room temperature.
separated by 3.71 A. Our earlier communication notes a
similar bimolecular stacking in symmetry-related mole-
62
˚
cules of 9 where ring-centroids are separated by 3.69 A.
Group 11 metal-ion complexes of azadipyrromethene
Such interactions appear in other structures, and there is
no apparent correlation with ring-substitution of the
azadipyrromethene. NMR evidence is consistent with
persistence of the intramolecular π-stacking interactions
in solution.71 Resonances associated with the “proximal”
azadipyrromethene phenyl rings, which undergo π-stack-
ing with PPh3 in the solid state, show a pronounced 0.3-
0.7 ppm upfield shift at room temperature in CDCl3
solution, whereas the same resonances in the PEt3-bound
complexes only shift by ∼0.1 ppm.
ligands substantially retain the optical properties of their
boron analogues. Figure 3 depicts absorption spectra of
homologous complexes 1, 4, and 9, together with free La;
Table 4 summarizes absorption maxima of all com-
pounds. A near-UV feature is present in all of the ligands
and complexes; this feature red-shifts for ligands bearing
a methoxy substituent on the proximal phenyl ring
(complexes 2 and 5, Table 4). This reddening of the
higher-energy absorption also takes place in zinc bis(aza-
dipyrromethenes). It originates from proximal phenyl-
group participation in the ligand-centered transitions that
elicit this near-UV feature.60 An intense visible absorp-
tion in the red region of the spectrum (ε ∼ 30 000 M-1
cm-1) is characteristic of azadipyrromethene ligands; a
shoulder is evident near 550 nm. Upon complexation this
absorption intensifies markedly; for silver complex 4, the
extinction coefficient at λmax more than doubles (ε = 65
000 M-1 cm-1). The amplification of the molar absorp-
tivity is not as dramatic for CuI and AuI complexes, and is
also less pronounced for PEt3AgI complexes 7 and 8. For
CuI complexes 1-3, the position of the visible absorption
maximum is nearly identical to that of the free azadipyr-
romethene. For AgI complexes 4-8 , the visible absorp-
tion maximum bathochromically shifts; the largest such
shift is 18 nm in 4 relative to La. For AuI complex 9, the
visible absorption maximum red-shifts by 3 nm from that
of tetraphenylazadipyrromethene. Methoxy-substituted
Optical Properties. Recent investigations of boron
azadipyrromethenes have emphasized light absorption
and emission. A persistent context has been photody-
namic therapy. Clinical utility of boron azadipyrro-
methenes or other triplet-state photosensitizers requires
absorptivity between about 600-900 nm. Light of these
wavelengths effectively penetrates human flesh; the long-
er wavelength limit is fixed by the excitation energy of
promoting 3O2 to its singlet excited state for therapeutic
action. Boron azadipyrromethenes absorb strongly be-
tween 650-700 nm with molar extinction coefficients in
CHCl3 of 75 000 M-1 cm-1 or greater. A corresponding
emission profile mirrors this absorption. Typical Stokes
shifts are 20-40 nm red of the absorption maximum.45
(71) Martin, C. B.; Mulla, H. R.; Willis, P. G.; Cammers-Goodwin, A. J.
Org. Chem. 1999, 64, 7802–7806.

8140 Inorganic Chemistry, Vol. 48, No. 17, 2009
Teets et al.
Table 5. Summary of Emission Maxima and Quantum Yields for All Complexesa
λem (nm)
Φem
La
Lb
Lc
1
2
3
642
662
660, 707 (sh)
642
665
667, 705 (sh)
647
667
0.0014
0.014
0.0020
0.0025
0.0067
0.0024
0.0039
0.012
4
5
6
7
8
9
660, 705 (sh)
0.0034
0.0021
0.0030
0.0024
645
642
645
a Measured in optically dilute deoxygenated CHCl3 solutions at room
temperature with 550 nm excitation.
Figure 4. Normalized room temperature emission spectra for 4 (red), 5
(green), and 6 (blue), recorded at room temperature in deoxygenated
CHCl3 solution at micromolar concentrations with 550 nm excitation.
effect. Also, the distinct shoulder that is evident in the
spectrum of 6 appears in the spectra of azadipyrro-
methene ligand Lc and its other complexes 3 and 8. The
room temperature emission spectra of the complexes are
similar to those of the free ligands, and are not severely
altered by changing the metal. In addition, replacing PPh3
in 4 and 6 with PEt3 in 7 and 8 does not modulate the
luminescence features. Emission quantum yields are also
very similar to those of the free azadipyrromethenes.
PPh3AgL b (5), Φem = 0.012, was the only metal complex
with a quantum yield exceeding 1%. Table 5 summarizes
the emission wavelengths and quantum yields for all
ligands and complexes at room temperature. As discussed
previously,62 emission spectra of 1, 4, and 9 measured at
77 K in 2-methyltetrahydrofuran glass display structured
luminescence, with average peak-to-peak separations
consistent with azadipyrromethene skeletal modes.72
The inefficient emission found at room temperature
may result from the rotational freedom of the four
pendant benzene rings, and possibly from the ancillary
phosphine ligands; conformationally restrained azadi-
pyrromethenes fluoresce with higher yields.73,74
ligands and complexes typically show red-shifted visible
absorption maxima compared to their unsubstituted
analogues; this perturbation in the absorption spectrum
is considerably more pronounced when the methoxy
group is positioned on the proximal phenyl ring (i.e.,
the phenyl substituent derived from acetophenone), as in
Lb and complexes 2 and 5. The intensity of the visible
absorption bands is not affected by the presence or
absence of methoxy groups on the pendant phenyl rings.
As described earlier,62 at 77 K in 2-methyltetrahydro-
furan glass the absorption features are much better re-
solved, and at least three distinct maxima in the red region
of the spectrum are evident for complexes 1, 4, and 9.
Absorption spectra of the new complexes bear simila-
rities to those of our earlier zinc(II) and mercury(II)
bis(azadipyrromethenes)60 and to the M(II) complexes
of O’Shea and collaborators.61 In bis(azadipyrro-
methene) complexes, visible absorption maxima approxi-
mately coincide with that of the free ligand, and with
maxima of the new complexes reported here. Molar
absorptivities at λmax are roughly twice those of the free
ligand. Group 11 mono(azadipyrromethenes) show a
single asymmetric visible absorption band, whereas bis-
(azadipyrromethene) complexes of 3d-metals (but not
mercury) show a partial splitting that is clearly resolved
for CuII(La)2. The near-ultraviolet absorption maximum
is similar across boron complexes and mono- and bis-
(azadipyrromethene) complexes. Substituent effects on
these higher-energy transitions were discussed earlier.60
Azadipyrromethene complexes of the d10 metal ions of
group 11 are luminescent. Room-temperature excitation
in chloroform elicits weak emission that mirrors the
absorption profile near 600 nm, with average Stokes shifts
of 41 nm for 1-9. The small Stokes shift suggests a
fluorescence origin of the emission. Excitation spectra
parallel absorption spectra. Figure 4 depicts normalized
room temperature emission spectra of AgI complexes 4-
6, which are representative of the complexes described
here. Absorption and emission spectra for complexes 1 -
9 are deposited as Supporting Information (Figures S1-
S9); optical spectra of 1, 4, and 9 were previously com-
municated.62 As shown in Figure 4, methoxy substitution
leads to a red-shift in the emission maximum, and again
substitution at the proximal phenyl ring has the greater
Calculations. DFT calculations have been performed
on the neutral ligand La (where one pyrrole nitrogen
is protonated) and on the model complexes (H3P)MILa
(M = Cu, Ag, Au). Table S1 (Supporting Information)
collects calculated interatomic distances alongside aver-
age values found crystallographically for 1, 4, and 9.
Root-mean square and average deviations are indicated.
Save for gold complex 9, optimized metrics compare
favorably with crystallographic values. In 9, microsym-
metry about gold is essentially C2v with a small difference
˚
(0.01 A) in the Au-N bond lengths. The calculated
structure predicts asymmetric gold ligation, with Au-N
bond lengths that differ by 0.385 A. Gold(I) in the
˚
optimized structure is pseudo-two coordinate with one
P-Au-N angle calculated at 162.3ꢀ. The more distant
pyrrole nitrogen perturbs the binding of gold from an
idealized linearity, but the calculation fails to reproduce
the observed trigonal coordination. Geometry optimiza-
tion of a model complex where Ph3PAuþ coordinates a
(72) Li, X.-Y.; Czernuszewicz, R. S.; Kincaid, J. R.; Su, Y. O.; Spiro, T. G.
J. Phys. Chem. 1990, 94, 31–47.
(73) Zhao, W.; Carreira, E. M. Angew. Chem., Int. Ed. 2005, 44, 1677–
1679.
(74) Loudet, A.; Bandichhor, R.; Burgess, K.; Palma, A.; McDonnell, S.
O.; Hall, M. J.; O’Shea, D. F. Org. Lett. 2008, 10, 4771–4774.

Article
Inorganic Chemistry, Vol. 48, No. 17, 2009 8141
variant of La where hydrogen replaces the remote phenyl
groups converges to a similar quasi-linear coordination
geometry. The cause of three-coordination of gold in 9
remains unresolved. The possibility that crystal-packing
interactions determine the solid-state structure cannot be
excluded, though the solution 1H NMR spectrum is
consistent with a Cs-symmetric structure. Discussion
henceforth will consider only 1 and 4, along with free La
and its BF2þ adduct.
Time-dependent DFT computations were performed
using the statistical averaging of (model) orbital poten-
tials (SAOP)75-77 on geometries optimized with the
revised Perdew-Becke-Ernzerhof functional.78,79 Re-
sponse calculations include implicit solvation through
the conductor-like screening model of Klamt and co-
workers.80-82 Vertical excitation wavelengths, oscillator
strengths, and orbital compositions of the excited states
were calculated.
In boron azadipyrromethenes, the optical transitions of
interest to photodynamic therapy center near 620 nm with
extinction coefficients reaching past 70 000 M-1 cm-1
.
Similar absorptions recur in metallocomplexes and in the
free ligand with lesser intensity. Figure 5 depicts a partial
Kohn-Sham orbital energy-level diagram of the model
copper complex PH3CuLa; plots of selected orbitals
appear as Figure 5(b). Time-dependent DFT calculations
indicate that multiple electronic transitions combine to
produce these absorptions. For all compounds, the lowest
several allowed transitions are to the lowest-occupied
Kohn-Sham orbital (LUMO), consistent with the
large one-electron energy separation between it and the
LUMO þ 1. In the free ligand, the LUMOrHOMO and
LUMOrHOMO - 1 transitions intermix through con-
figuration interaction. Much the same arrangement pre-
vails in BF2þ and group 11 chelates. For example, in the
model copper(I) complex PH3CuLa HOMOrLUMO
and HOMOrLUMO - 3 transitions undergo configura-
tion interaction and yield vertical excitations of high
oscillator strength. The orbitals involved are entirely
analogous to those of the free ligand with minimal
participation of copper.
The visible absorption maxima of azadipyrromethene
complexes red-shift with para-methoxy substitution of
either the proximal or the distal phenyl substituents, cf.,
Table 4. This effect has been noted previously,45 and is
more pronounced for substitution at the proximal (to
copper) phenyls. Time-dependent DFT calculations cap-
ture the red-shifting effect of methoxy substitution. The
transitions with the highest oscillator strength move 18
nm to lower energy for PH3CuLb relative to the complex
of La, and a lesser red-shift is calculated for PH3CuLc.
Figure 5. (a) Partial Kohn-Sham orbital eigenvalue diagram of model
complexes PH3CuLa-c. Implicit chloroform solvation is includedthrough
a conductor-like screening model. (b) Plots of selected orbitals of
(H3P)CuLa (contour level 0.03 au).
The energy-level diagram of Figure 5 suggests an
explanation. Methoxy substitution destabilizes the
HOMO and LUMO of Lb and Lc complexes, compared
to those of La; the HOMO rises slightly more in energy
than the LUMO.83 Thus, para-methoxy substitution
shrinks the HOMO-LUMO gap, in a manner consistent
with the red-shifted absorptions found experimentally.
The orbital plots of Figure 5(b) show amplitude on
the para carbon atoms of the HOMO and LUMO of
PH3CuLa, suggesting that both orbitals respond to sub-
stitution at these positions. The greater orbital amplitude
rests at the para carbon of the proximal phenyl rings, and
substitution here is more red-shifting than at the distal
centers.
Azadipyrromethene electronic transitions are interpre-
table within a three-orbital model, Scheme 3. As in the
spectra of porphyrin species,84-86 the most prominent
features result from transitions undergoing configuration
interaction. In Scheme 3, state designations refer to the
C2v-symmetric boron complex 10. Time-dependent cal-
culation of 10 predicts a pair of intense absorptions, each
having the same two constituent excitations. These are the
(75) Schipper, P. R. T.; Gritsenko, O. V.; van Gisbergen, S. J. A.;
Baerends, E. J. J. Chem. Phys. 2000, 112, 1344–1352.
(76) Gritsenko, O. V.; Schipper, P. R. T.; Baerends, E. J. Chem. Phys.
Lett. 1999, 302, 199–207.
::
(77) Gruning, M.; Gritsenko, O. V.; van Gisbergen, S. J. A.; Baerends, E.
J. J. Chem. Phys. 2002, 116, 9591–9601.
(78) Perdew, J. P.; Burke, K.; Ernzerhof, M. Phys. Rev. Lett. 1996, 77,
3865–3868.
(79) Hammer, B.; Hansen, L. B.; Norskov, J. K. Phys. Rev. B 1999, 59,
(83) Abbreviations: HOMO, highest occupied Kohn-Sham orbital;
LUMO, lowest unoccupied Kohn-Sham orbital.
(84) Gouterman, M. Optical spectra and electronic structure of porphyr-
ins and related rings. In The Porphyrins; Dolphin, D., Ed.; Academic Press:
New York, 1978; pp 1-165
7413–7421.
:: ::
(85) Loh, Z.-H.; Miller, S. E.; Chang, C. J.; Carpenter, S. D.; Nocera,
D. G. J. Phys. Chem. A 2002, 106, 11700–11708.
(86) Partyka, D. V.; Robilotto, T. J.; Zeller, M.; Hunter, A. D.; Gray,
T. G. Proc. Natl. Acad. Sci., U.S.A. 2008, 105, 14293–14297.
(80) Klamt, A.; Schuurmann, G. J. Chem. Soc., Perkin Trans. 2 1993,
799–805.
(81) Klamt, A. J. Phys. Chem. 1995, 99, 2224–2235.
(82) Klamt, A.; Jones, V. J. Chem. Phys. 1996, 105, 9972–9981.

8142 Inorganic Chemistry, Vol. 48, No. 17, 2009
Teets et al.
Scheme 3. Configuration Interaction in the Absorption Spectra of
transition dipole moments, and not by relative weights in
configuration interaction.
Azadipyrromethene Complexes Depicted for the Parent BF2þ Complex
a
(C2v
)
Conclusions
Syntheses and characterizations of three-coordinate azadi-
pyrromethene complexes with group 11 metals are described.
The copper(I) and silver(I) compounds are prepared by room
temperature reactions of free azadipyrromethene ligands
with (phosphine)metal(I) triflates in the presence of an amine
base. We find that the synthetic methods described here are
quite general, allowing for methoxy-substituted tetraaryla-
zadipyrromethene complexes to be prepared, and permitting
the facile substitution of triethylphosphine for triphenylpho-
sphine. Ligand 1H NMR resonances are generally unrespon-
sive to metal complexation, though resonances associated
with the proximal aryl rings do shift upfield by more than
0.3 ppm when PPh3 occupies the third coordination site.
NMR methods, in combination with combustion analysis,
have established the purity of all new compounds, which are
isolated in 65% or better yields. The structural features of the
new complexes described here recall previously reported
complexes. All complexes are tricoordinate with slight asym-
metries in M-N bond lengths. The N-M-N coordination
plane is canted relative to the mean plane of the ligand
chromophore, and this canting is not relieved upon substitu-
tion of less sterically demanding PEt3 for PPh3. Also, both
intra- and intermolecular π-stacking interactions, involving
PPh3 and pendant azadipyrromethene phenyls, are fre-
quently observed in the solid state. 1H NMR spectra suggest
that these interactions persist in solution as well. All metal
complexes described retain intense electronic absorption
features in the red region of the visible spectrum, and weak
red emission occurs. Methoxy-substituted azadipyrro-
methene complexes demonstrate red-shifted absorption and
emission maxima relative to unsubstituted complexes, whicþh
mirrors the trend observed for previously characterized BF2
and Zn(II) complexes.44,45,60 At 77 K, the resolution of
absorption and emission spectra is markedly improved,
allowing for the discernment of at least three distinct absorp-
tion bands in the red region, and regular structure in the
emission spectra.62 The optical spectra suggest ligand-based
electronic transitions, which time-dependent DFT calcula-
tions corroborate.
a Orbital contour levels are 0.03 a.u.
LUMOrHOMO and LUMOrHOMO - 1 transitions,
or in the notation of Scheme 3, b1r2a2 and b1r1a2.
The singlet states so derived are both B2, and they
engage in configuration interaction. The b1r2a2
(LUMOrHOMO) excitation carries a larger transition
moment than the b1r1a2 (LUMOrHOMO - 1) promo-
tion. For unsubstituted 10, the two transitions contribute
roughly equally to both absorptions. In the lower-energy
state, the transition dipole vectors are opposed.
Time-dependent calculations of tetraphenyl azadipyr-
romethene (La) complexes find that analogous transitions
dominate the intense absorption calculated at 601 nm. As
with 10, transition dipole vectors for excitation to the
lowest singlet excited state of (H3P)CuLa oppose each
other, with a 168ꢀ angle between them. However, the
b1r2a2 (LUMOrHOMO) component carries a greater
transition dipole moment than its higher-energy CI coun-
terpart b1r1a2 (here, LUMOrHOMO - 3). So much
greater is the norm of the LUMOrHOMO transition
dipole vector that it predominates, and the 601 nm
transition, primarily LUMOrHOMO, is more intense.
Its companion transition, at 538 nm, is mainly b1r1a2,
and the dipole contribution from b1r2a2 is augmenta-
tive. Transition dipole vectors for the 538-nm excitation
lie at 11.6ꢀ to each other in PH3CuLa. However, the
smaller b1r1a2 dipole vector leaves this additive transi-
tion with a lesser oscillator strength. Thus, an intense
transition is calculated at 601 nm for PH3CuLa with a
minor absorption at 538 nm, corresponding to the
b1r2a2 and b1r1a2 excitations of Scheme 3. Calcula-
tions for PH3AgLa and for BF2-azadipyrromethene 11
are broadly similar. Absorption profiles of azadipyrro-
methene fluorophores within the phototherapeutic win-
dow (ca. 600-900 nm) are determined by constitutive
The experiments herein, along with our earlier report60 of
group 12 bis(azadipyrromethenes), and O’Shea and collabora-
tors’ disclosure61 of first-row transition element azadipyrro-
methene complexes attest that these ligands are versatile. The
visible absorption of azadipyrromethene ligands has for some
years44 commended their boron complexes for photodynamic
therapy. The distinctive optical profiles of azadipyrromethenes
persist upon complexing main group60 and d-block metal
centers.61 Calculations described here show that the visible
and near-ultraviolet photophysics of group 11 azadipyrro-
methenes are ligand-centered. The absorption properties are
reminiscent of porphyrin electronic spectra in that two one-
electron promotions to states of the same symmetry mingle
through configuration interaction. The chelated Lewis acid
minimally influences absorption or emission maxima. Optical
transitions are more sensitive to methoxy substitution of the
proximal (Lb series) or distal (Lc series) phenyl substituents.
This sensitivity results from frontier orbital amplitude at the
para-carbon atoms of all four benzene rings.

Article
Investigations of phototherapy and cellular labeling with
Inorganic Chemistry, Vol. 48, No. 17, 2009 8143
stirring for 1 h at room temperature, the solvent was stripped via
rotary evaporation. The resulting residue was taken up in 20 mL
of benzene, and washed with 2 ꢀ 20 mL of H2O. The organic
layer was dried with MgSO4 and evaporated to dryness. The
product was crystallized from a concentrated THF solution via
room temperature vapor diffusion of pentane. Yield: 58 mg
(67%). 1H NMR (CDCl3): δ 7.95-7.98 (m, 8H), 7.29-7.35 (m,
9H), 7.18-7.22 (m, 8H), 6.90-6.95 (m, 6H), 6.35 (d, 4H), 3.52 (s,
6H). 31P NMR (CDCl3): δ 15.2 (d, br) ppm. UV-vis (CHCl3): λ
(ε) 324 (30000), 631 (69000) nm. Emission (CHCl3, λex = 550
nm): λ 667 nm. Anal. Calcd for C52H41N3O2PAg: C, 71.07; H,
4.70; N, 4.78. Found: C, 70.53; H, 4.40; N, 4.66.
azadipyrromethene sensitizers will doubtless continue, but
emerging ideas in medicinal chemistry pose new opportu-
nities. Azadipyrromethene complexes of paramagnets such
as gadolinium(III), or of radionuclides such as 99mTc or 111In
hold potential as dual-mode optical/MRI or optical/radio-
imaging agents, respectively. Their clinical applications, if
any, lie in the future, but exciting discoveries with azadipyrro-
methenes are surely at hand.
Experimental Section
PPh3AgLc (6). Prepared in the same manner as for 5, using 51
mg (0.098 mmol) of PPh3AgOTf and 50 mg (0.098 mmol) of Lc;
all other conditions identical to above. Yield: 59 mg (68%). 1H
NMR (CDCl3): δ 8.03 (dd, 4H), 7.94 (d, 4H), 7.32 (td, 3H),
7.16-7.20 (m, 8H), 6.83-6.92 (m, 16H), 3.88 (s, 6H). 31P NMR
(CDCl3): δ 13.2 (s, br) ppm. UV-vis (CHCl3): λ (ε) 304 (39000),
615 (64000) nm. Emission (CHCl3, λex = 550 nm): λ 660, 705
(sh). Anal. Calcd for C52H41N3O2PAg: C, 71.07; H, 4.70; N,
4.78. Found: C, 70.68; H, 4.31; N, 5.13.
General Procedures. Triphenylphosphine (PPh3), triethyl-
phosphine (PEt3), AgOTf (OTf = trifluoromethanesulfonate),
diisopropylethylamine (DIPEA), and (CuOTf)2 toluene were
3
obtained commercially and used as received. PPh3AgOTf87 and
tetraarylazadipyrromethene ligands La-c were prepared as pre-
viously described.45 The syntheses of 1, 4, and 9 were described
earlier.62 Air-sensitive solids were handled in an inert-atmo-
sphere glovebox operating at <10 ppm O2. Reactions involving
air-sensitive materials were carried out under an atmosphere of
Argon using standard Schlenk-line techniques; solvents for such
reactions were purified with an M-Braun SPS and purged with
argon prior to use. 1H and 31P NMR spectra were recorded on a
Varian AS-400 spectrometer, operating at 400 and 161.8 MHz,
respectively. Microanalyses were performed by Quantitative
Technologies Inc. and Robertson Microlit Laboratories.
Absorption and corrected emission spectra were recorded in
HPLC-grade chloroform on a Cary 5G spectrophotometer and
a Cary Eclipse fluorimeter, respectively. Fluorescence quantum
yields were determined using deoxygenated, optically dilute
solutions with absorbance <0.1 at the respective excitation
wavelength. (Bu4N)2Mo6Cl14 was used as a standard with a
literature value for Φf of 0.19.88
PEt3AgLa (7). PEt3AgOTf was generated in situ by stirring
AgOTf (26 mg, 0.10 mmol) and PEt3 (22 μL, 0.15 mmol) in 5 mL
of THF at room temperature for 20 min. To the resulting
colorless solution was added a solution of La (43 mg, 0.096
mmol) in 7 mL of THF via syringe. Finally, DIPEA (0.16 mL,
0.98 mmol) was added via syringe. After stirring for 1 h at room
temperature, the solvent was removed via rotary evaporation.
The resulting residue was taken up in 30 mL of benzene, and
washed with 2 ꢀ 20 mL of H2O. The organic layer was dried over
MgSO4 and evaporated to give a brown solid. The product was
recrystallized from a concentrated THF solution by room
temperature vapor diffusion of pentane. Yield: 42 mg (65%).
1H NMR (CDCl3): δ 8.03 (d, 4H), 7.96 (d, 4H) 7.30-7.45 (m,
12H), 7.19 (s, 2H), 1.05-1.13 (m, 6H), 0.63-0.71 (m, 9H). 31
P
PPh3CuLb (2). PPh3CuOTf was generated in situ by stirring
(CuOTF)2 toluene (26 mg, 0.050 mmol) and PPh3 (27 mg,
NMR (CDCl3): δ 8.2 (s, br) ppm. UV-vis (CHCl3): λ (ε) 304
(30000), 610 (42000) nm. Emission (CHCl3, λex = 550 nm): λ,
645 nm. Anal. Calcd for C38H37N3PAg: C, 67.66; H, 5.53; N,
6.23. Found: C, 68.25; H, 5.53; N, 6.23.
3
0.10 mmol) in 5 mL of THF for 20 min at room temperature.
To the resulting solution was added a solution of Lb (53 mg,
0.10 mmol) in 7 mL of THF via syringe, followed by DIPEA
(0.16 mL, 0.98 mmol). The reaction mixture was stirred at room
temperature for 1 h, at which time all volatiles were removed in
vacuo. The resulting residue was dried in vacuo overnight to give
a bluish-black solid, which was taken up in about 4 mL of
toluene, filtered through Celite, and crystallized by vapor diffu-
sion of pentane in a nitrogen glovebox. Yield: 66 mg (76%). 1H
NMR (CDCl3): δ 8.04 (d, 4H), 7.88 (d, 4H), 7.26-7.39 (m, 9H),
PEt3AgLc (8). Prepared in same manner as for 7, using 26 mg
(0.10 mmol) of AgOTf, 22 μL (0.15 mmol) of PEt3, and 49 mg
(0.096 mmol) of Lc; all other conditions identical to above.
Yield: 49 mg (69%). 1H NMR (CDCl3): δ 8.02 (d, 4H), 7.93 (d,
4H), 7.33-7.44 (m, 6H), 7.11 (s, 2H), 6.90 (d, 4H), 3.87 (s, 6H),
1.05-1.11 (m, 6H), 0.62-0.70 (m, 9H). 31P NMR (CDCl3): δ 8.8
(d, br) ppm. UV-vis (CHCl3): λ (ε) 303 (31000), 614 (46000) nm.
Emission (CHCl3, λex = 550 nm): λ 661, 702 (sh) nm. Anal.
Calcd For C40H41N3O2PAg: C, 65.40; H, 5.62; N, 5.72. Found:
C, 66.03; H, 5.69; N, 5.73 .
7.15-7.19 (m, 8H), 6.93 (t, 6H), 6.42 (d, 4H), 3.60 (s, 6H). 31
P
NMR (CDCl3): δ 3.5 (s, br) ppm. UV-vis (CHCl3): λ (ε) 320
(24000), 620 (28000) nm. Emission (CHCl3, λex = 550 nm): λ 665
nm. Anal. Calcd for C52H41N3O2PCu: C, 74.85; H, 4.95; N,
5.04. Found: C, 74.61; H, 4.69; N, 4.82.
X-ray Structure Determination. Products were crystallized by
diffusion of pentane into saturated toluene or THF solutions.
Single crystal X-ray data were collected on a Bruker AXS
SMART APEX CCD diffractometer using monochromatic
Mo KR radiation with omega scan technique. The unit cells
were determined using SMART89 and SAINTþ.90 Data collec-
tion for all crystals was conducted at 100 K (-173 ꢀC). All
structures were solved by direct methods and refined by full
matrix least-squares against F2 with all reflections using
SHELXTL.91 Refinement of extinction coefficients was found
to be insignificant. All non-hydrogen atoms were refined aniso-
tropically. All hydrogen atoms were placed in standard calcu-
lated positions and all hydrogen atoms were refined with an
PPh3CuLc (3). Prepared in the same manner as for 2, using
25 mg (0.048 mmol) of (CuOTF)2 toluene, 25 mg (0.095 mmol)
3
of PPh3, and 49 mg (0.096 mmol) of Lc1; all other conditions
identical to above. Yield: 63 mg (79%). H NMR (CDCl3): δ
8.01 (d, 4H), 7.91 (d, 4H), 7.26 (t, 3H), 7.11-7.15 (m, 8H), 6.83-
6.98 (m, 16H), 3.89 (s, 6H). 31P NMR (CDCl3): δ 3.5 (s, br) ppm.
UV-vis (CHCl3): λ (ε) 306 (32000), 606 (38000) nm. Emission
(CHCl3, λex = 550 nm): λ, 667, 705(sh) nm. Anal. Calcd for
C52H41N3O2PCu: C, 74.85; H, 4.95; N, 5.04. Found: C, 74.73;
H, 4.85; N, 4.88.
PPh3AgLb (5). PPh3AgOTf (51 mg, 0.098 mmol) and Lb
(50 mg, 0.098 mmol) were dissolved in 12 mL of THF, and
DIPEA (0.16 mL, 0.98 mmol) was added via syringe. After
(89) Bruker Advanced X-ray Solutions, SMART for WNT/2000 (Version
5.628), Bruker AXS Inc., Madison, WI: USA, 1997-2002.
(90) Bruker Advanced X-ray Solutions, SAINT, Version 6.45; Bruker AXS
Inc.: Madison, WI, 1997-2003.
(91) Bruker Advanced X-ray Solutions, SHELXTL, Version 6.10; Bruker
AXS Inc.: Madison, WI, 2000.
(87) Lettko, L.; Wood, J. S.; Rausch, M. D. Inorg. Chim. Acta 2000, 308,
37–44.
(88) Maverick, A. W.; Najdzionek, J. S.; MacKenzie, D.; Nocera, D. G.;
Gray, H. B. J. Am. Chem. Soc. 1983, 105, 1878–1882.

8144 Inorganic Chemistry, Vol. 48, No. 17, 2009
Teets et al.
isotropic displacement parameter 1.2 times that of the adjacent
carbon.
tions verified that all converged structures are potential-energy
minima. Vertical excitation energies were calculated with single-
point time-dependent DFT calculations,103-105 performed with
the asymptotically correct LB94 functional of van Leeuwen
and Baerends106 or the statistical average of model exchange-
correlation potentials (SAOP)107 at the geometry previously
optimized with the Becke-Perdew functional. Basis sets for
time-dependent calculations are those used in geometry opti-
mization. Time-dependent calculations employed the Davidson
algorithm; error tolerances in the square of the excitation
energies and the trial-vector orthonormality cutoff were
both set to 10-8. Time-dependent calculations employed the
conductor-like screening model of Klamt and co-workers
(COSMO) using chloroform solvent,80-82 in single-point calcu-
lations on the gas-phase structures.
Computations. Scalar-relativistic DFT calculations were exe-
cuted within the zero-order regular approximation, as imple-
mented in the Amsterdam Density Functional package
(ADF2006).92-94 All calculations were spin-restricted. In these
computations, the local density functional is that of Vosko,
Wilk, and Nusair.95 Gradient corrections were introduced
with the exchange functional of Becke96 and the correlation
functional of Perdew.97,98 Scalar relativistic interactions
were included within the zero-order regular approximation
(ZORA).99-102 Atomic orbitals of all atoms were constructed
from triple-ζ basis sets of Slater orbitals (ADF database TZP).
Core electronic shells (one or more principle quantum numbers
below the valence level) were frozen for all atoms. The electron
density was fitted with sets of auxiliary s, p, d, f, and g-functions
centered on all nuclei, along with the Coulomb and exchange
potentials, in every self-consistent field cycle. Gas-phase geo-
metries were optimized with the Broyden-Fletcher-Gold-
farb-Shano algorithm; the convergence criterion was that
successive energies differ by not more than 1 ꢀ 10-4 H (the
default is 1 ꢀ 10-3 H). Numerical harmonic frequency calcula-
Acknowledgment. The authors thank the National Science
Foundation (Grant CHE-0749086 to T.G.G.). T.G.G. is an
Alfred P. Sloan Research Fellow. T.S.T. acknowledges the Case
Western Reserve University SOURCE program for funding; he
currently holds a John and Fannie Hertz Foundation predoc-
toral fellowship. The diffractometer at Case Western Reserve
was funded by NSF Grant CHE-0541766. We thank Professor
D. G. Nocera, Massachusetts Institute of Technology, for access
to instrumentation, and Dr. D. V. Partyka for experimental
assistance.
(92) SCM ; Theoretical Chemistry, Vrije Universiteit: Amsterdam, The
Netherlands, http://www.scm.com.
(93) te Velde, G.; Bickelhaupt, F. M.; Baerends, E. J.; Guerra, F. C.;
van Gisbergen, S. J. A.; Snijders, J. G.; Ziegler, T. J. Comput. Chem. 2001, 22,
931–967.
Supporting Information Available: X-ray crystallographic
data in CIF format, optical absorption and emission spectra
for ligands La-c and complexes 1-9. This material is available
free of charge via the Internet at http://pubs.acs.org.
(94) Guerra, F. C.; Visser, O.; Snijders, J. G.; te Velde, G.; Baerends, E. J.
In Methods and Techniques for Computational Chemistry; Clementi, E.,
Corongiu, C., Eds.; STEF: Cagliari, 1995; pp 303-395.
(95) Vosko, S. H.; Wilk, L.; Nusair, M. Can. J. Phys. 1980, 58, 1200–1211.
(96) Becke, A. D. Phys. Rev. A 1988, 38, 3098–3100.
(97) Perdew, J. P.; Zunger, A. Phys. Rev. B 1981, 23, 5048–5079.
(98) Perdew, J. P. Phys. Rev. B 1986, 33, 8822–8824.
(99) Van Lenthe, E.; Baerends, E. J.; Snijders, J. G. J. Chem. Phys. 1993,
99, 4597–4610.
(100) Van Lenthe, E.; Baerends, E. J.; Snijders, J. G. J. Chem. Phys. 1994,
101, 9783–9792.
(101) Van Lenthe, E.; Snijders, J. G.; Baerends, E. J. J. Chem. Phys. 1996,
105, 6505–6516.
(103) van Gisbergen, S. J. A.; Kootstra, F.; Schipper, P. R. T.; Gritsenko,
O. V.; Snijders, J. G.; Baerends, E. J. Phys. Rev. A 1998, 57, 2556–2571.
(104) Jamorski, C.; Casida, M. E.; Salahub, D. R. J. Chem. Phys. 1996,
104, 5134–5147.
(105) Bauernschmitt, R.; Ahlrichs, R. Chem. Phys. Lett. 1996, 256, 454–
464.
(106) van Leeuwen, R.; Baerends, E. J. Phys. Rev. A 1994, 49, 2421–2431.
(107) Schipper, P. R. T.; Gritsenko, O. V.; van Gisbergen, S. J. A.;
Baerends, E. J. J. Chem. Phys. 2000, 112, 1344–1352.
(102) Van Lenthe, E.; van Leeuwen, R.; Baerends, E. J.; Snijders, J. G. Int.
J. Quantum Chem. 1996, 57, 281–293.