A paradigm for blue- or red-shifted absorption of small molecules depending on the site of π-extension

Article abstract of DOI:10.1021/ja1075162

Benzannulation of aromatic molecules is often used to red-shift absorption and emission bands of organic and inorganic, molecular, and polymeric materials; however, in some cases, either red or blue shifts are observed, depending on the site of benzannulation. A series of five platinum(II) complexes of the form (NNN)PtCl are reported here that illustrate this phenomenon, where NNN represents the tridentate monoanionic ligands 2,5-bis(2-pyridylimino)3,4- diethylpyrrolate (1), 1,3-bis(2-pyridylimino)isoindolate (2), 1,3-bis(2-pyridylimino)benz(f)isoindolate (3), 1,3-bis(2-pyridylimino)benz(e) isoindolate (4), and 1,3-bis(1-isoquinolylimino) isoindolate (5). For this series of molecules, either a blue shift (2 and 3) or a red shift (4 and 5) in absorption and emission maxima, relative to their respective nonbenzannulated compounds, was observed that depends on the site of benzannulation. Experimental data and first principles calculations suggest that a similar HOMO energy level and a destabilized or stabilized LUMO with benzannulation is responsible for the observed trends. A rationale for LUMO stabilization/destabilization is presented using simple molecular orbital theory. This explanation is expanded to describe other molecules with this unusual behavior.

Full text of DOI:10.1021/ja1075162

Published on Web 10/21/2010
A Paradigm for Blue- or Red-Shifted Absorption of Small
Molecules Depending on the Site of π-Extension
Kenneth Hanson,† Luke Roskop,‡ Peter I. Djurovich,† Federico Zahariev,‡
Mark S. Gordon,*,‡ and Mark E. Thompson*,†
Department of Chemistry, UniVersity of Southern California, Los Angeles, California 90089, United
States, and Department of Chemistry, Iowa State UniVersity, Ames, Iowa 50011, United States
Received August 19, 2010; E-mail: mark@si.msg.chem.iastate.edu; met@usc.edu
Abstract: Benzannulation of aromatic molecules is often used to red-shift absorption and emission bands
of organic and inorganic, molecular, and polymeric materials; however, in some cases, either red or blue
shifts are observed, depending on the site of benzannulation. A series of five platinum(II) complexes of the
form (N∧N∧N)PtCl are reported here that illustrate this phenomenon, where N∧N∧N represents the tridentate
monoanionic ligands 2,5-bis(2-pyridylimino)3,4-diethylpyrrolate (1), 1,3-bis(2-pyridylimino)isoindolate (2),
1,3-bis(2-pyridylimino)benz(f)isoindolate (3), 1,3-bis(2-pyridylimino)benz(e)isoindolate (4), and 1,3-bis(1-
isoquinolylimino) isoindolate (5). For this series of molecules, either a blue shift (2 and 3) or a red shift (4
and 5) in absorption and emission maxima, relative to their respective nonbenzannulated compounds,
was observed that depends on the site of benzannulation. Experimental data and first principles calculations
suggest that a similar HOMO energy level and a destabilized or stabilized LUMO with benzannulation is
responsible for the observed trends. A rationale for LUMO stabilization/destabilization is presented using
simple molecular orbital theory. This explanation is expanded to describe other molecules with this unusual
behavior.
Introduction
to this report is the red shift observed when going from benzene
to naphthalene to anthracene.4 However, there are several
The ability to readily tune the optical absorption properties
of molecular materials is important for a large number of
applications ranging from solar energy conversion to photody-
namic therapy.1 The absorption properties of a molecule or
ligand can be controlled by structural modification of the π-
conjugated system(s). Two common approaches to tailor the
photophysical properties of a molecule, without changing the
parent structure, are substitution with donor and/or acceptor
groups or extending π-conjugation. The addition of electron-
donating and/or electron-withdrawing groups can be used to
induce either hypsochromic (blue) or bathochromic (red) shifts
in the absorption/emission spectra, by shifting the HOMO or
LUMO orbital relative to the unsubstituted compound.2 On the
other hand, extending conjugation, particularly through ben-
zannulation of aromatic rings, is commonly assumed to desta-
bilize the highest occupied molecular orbital (HOMO) and
stabilize the lowest unoccupied molecular orbital (LUMO). This
decrease in separation between the HOMO and LUMO energies
will inevitably lead to red-shifted spectra.3 An example relevant
examples of small molecules that undergo a blue shift in
absorption upon benzannulation.5 This phenomenon was most
recently reported for benzannulation of the square planar 1,3-
bis(2-pyridylimino)isoindolate platinum chloride, (BPI)PtCl,
complex (2 in Chart 1).6 Although rationalizations for the
unexpected blue shifts have accompanied these reports, a
generalized description with predictive capability to explain the
effect of benzannulation on absorption has not been given.
To study the effect of benzannulation on the absorption of
small molecules, we have focused on a series of 1,3-bis(2-
pyridylimino)isoindolate platinum chloride derivatives. These
complexes are thermally stable, readily synthesized,7 easily
modified, emissive at room temperature, and have high molar
absorptivities.6
Herein, we present the synthesis and characterization of a
family of (BPI)PtCl derivatives (Chart 1) for which either red
or blue shifts are observed when the π-system is extended, for
example, λmax(absorption) for complexes 2, 3, and 4 are 487,
477, and 529, respectively. The direction of the shift upon
† University of Southern California.
(4) Turro, N. J.; Ramamurthy, V.; Scaiano, J. C. Principles of Molecular
Photochemistry: An Introduction; University Science Books: Sausalito,
CA, 2009; pp 30-31.
‡ Iowa State University.
(1) (a) Acc. Chem. Res. 2009, 42, 1859 (special issue). (b) Celli, J. P.;
Spring, B. Q.; Rizvi, I.; Evans, C. L.; Samkoe, K. S.; Verma, S.; Pogue,
B. W.; Hasan, T. Chem. ReV. 2010, 110, 2795–2838.
(2) Zollinger, H. Color Chemistry, 3rd ed.; Wiley-VCH: Zurich, 2001; p
550.
(3) (a) Griffiths, J. Colour and Constitution of Organic Molecules;
Academic Press: New York, 1976; p 281. (b) Rogers, J. E.; Nguyen,
K. A.; Hufnagle, D. C.; McLean, D. G.; Su, W.; Gossett, K. M.; Burke,
A. R.; Vinogradov, S. A.; Pachter, R.; Fleitz, P. A. J. Phys. Chem. A
2003, 107, 11331–11339.
(5) (a) Juris, A.; Barigelletti, F.; Balzani, V.; Belser, P.; Von Zelewsky,
A. Inorg. Chem. 1985, 24, 202–206. (b) Naraso; Nishida, J.-i.; Ando,
S.; Yamaguchi, J.; Itaka, K.; Koinuma, H.; Tada, H.; Tokito, S.;
Yamashita, Y. J. Am. Chem. Soc. 2005, 127, 10142–10143. (c) Adachi,
M.; Murata, Y. J. Phys. Chem. A 1998, 102, 841–845. (d) Adachi,
M.; Nagao, Y. Chem. Mater. 2001, 13, 662–669.
(6) Wen, H.-M.; Wu, Y.-H.; Fan, Y.; Zhang, L.-Y.; Chen, C.-N.; Chen,
Z.-N. Inorg. Chem. 2010, 49, 2210–2221.
(7) Siegl, W. O. J. Org. Chem. 1977, 42, 1872–1878.
9
10.1021/ja1075162  2010 American Chemical Society
J. AM. CHEM. SOC. 2010, 132, 16247–16255 16247

A R T I C L E S
Hanson et al.
Chart 1
cyclooctadiene (Aldrich), 1,2-dicyanonaphthalene, 2,3-dicyanon-
aphthalene (TCI America), calcium chloride (J.T. Baker), and
K2PtCl4 (Pressure Chemical Co.) were purchased from the corre-
sponding supplier (in parentheses) and used without further
purification for synthesis. (COD)PtCl2,10 (E)-3,4-dibromo-3-hex-
ene,11 (E)- and (Z)-3,4-dicyano-3-hexene,12 1,3-bis(2-pyridylimi-
no)isoindole,7 1,3-bis(2-pyridylimino)isoindolate platinum(II) chlo-
ride,13 and 1,3-bis(2-pyridylimino)benz[f]isoindole14 were synthesized
according to literature procedures. All reported NMR spectra were
obtained on a Bruker AC-250 MHz FT NMR or a Varian 400 MHz
NMR with all shifts given relative to residual solvent signals. Solid
probe mass spectrometer (MS) spectra were taken with a Hewlett-
Packard MS instrument with electron impact ionization and model
5973 mass selective detector. High resolution mass spectroscopy
was performed at the UCR Mass Spectrometry Facility at the
University of California, Riverside, CA. Elemental analyses (CHN)
were performed at the Microanalysis Laboratory at the University
of Illinois, Urbana-Champaign, IL.
benzannulation is dependent on the site of π-extension. Our
experimental and theoretical studies show that the spectral shifts
are dependent on the stabilization or destabilization of specific
orbitals (e.g., the LUMO) upon benzannulation. In other metal
complexes and organic molecules, similar red and blue shifts
have been observed as a result of site selective benzannulation.
Blue shifts in these compounds have been ascribed to molecular
distortions,8 shifts in total antibonding character,5c,d or increased
localization of π* orbitals.5a None of these explanations provide
a clear understanding of the origin of the shift or how to use it
to intentionally tune the transition energies in these materials.
The model presented here accurately predicts the direction of
the energy shift, due to benzannulation, of a wide range of
compounds and gives insight into the origin of the energy shift.
Quantum chemistry calculations are a common tool used to
interpret chemical behavior. In many cases, they can provide a
rationale that correlates nicely with the experiment, but the use
of molecular orbital interpretations is not always justified. The
time-dependent density functional theory (TDDFT), which is
typically used to calculate the excitations of larger systems,
complicates the molecular orbital analysis due to the fact that
a given TDDFT excitation could be composed of several
molecular-orbital transitions up in energy (“excitations”) as well
as several transitions down in energy (“de-excitations”). This
article introduces a scheme that one can routinely use to
determine the validity of an orbital argument based on the
Tamm-Dancoff approximation (TDA) to TDDFT, hereinafter
referred to as TDA-TDDFT, in which case the molecular
excitations are composed of molecular-orbital transitions only
up in energy (“excitations”).9 In addition, a dominant molecular-
orbital “excitation” may be selected by analyzing the transition
amplitudes. Then, further comparisons may be made between
the TDA-TDDFT excitation energy and the energy difference
corresponding to the dominant transition, for example, HOMOf
LUMO if the dominant transition is from the HOMO to the
LUMO, as well as with the actual experimental excitation. It
will be further shown that the molecules investigated in the
current work satisfy the conditions and therefore warrant the
use of molecular orbital interpretations.
2,5-Bis(2-pyridylimino)3,4-diethylpyrrole (BPEP). A mixture
of 2 g (14.92 mmol) of (Z)-3,4-dicyano-3-hexene, 2.8 g (29.8 mmol)
of 2-aminopyridine, and 0.4 g (4 mmol) of CaCl2 in 80 mL of
1-butanol was refluxed under N2 for 10 days. After removal of
1-butanol under reduced pressure, the remaining residue was
separated on a column of silica gel, eluting first with CH2Cl2, then
a mixture of CH2Cl2:ethyl acetate (9:1). The orange fraction was
collected and rotary evaporated to dryness to give 187 mg (4%) of
1
a orange solid. H NMR (250 MHz, CDCl3, δ): 8.52 (ddd, J ) 5,
2, 0.75 Hz, 2H), 7.74 (dd, J ) 7.5, 2 Hz, 2H), 7.38 (ddd, J ) 7.75,
1.25, 0.75 Hz, 2H), 7.09 (ddd, J ) 7.5, 5, 1.25 Hz, 2H), 2.6 (q, J
) 7.25 Hz, 4H), 1.24 (t, J ) 7.25 Hz, 6H).
2,5-Bis(2-pyridylimino)3,4-diethylpyrrolate platinum(II) Chlo-
ride (1). Quantities of 132 mg (0.354 mmol) of (COD)PtCl2 and
100 mg (0.328 mmol) of BPEP were suspended in 7 mL of
methanol. After the addition of 0.05 mL (0.354 mmol) of NEt3,
the mixture was heated to 60 °C under nitrogen overnight. The
mixture was then rotary evaporated to dryness and passed through
a plug of silica eluting with CH2Cl2. The product was then
recrystallized by dissolving the residue in a minimum amount of
CH2Cl2 and layering with methanol. The precipitate was collected
by filtration and washed with methanol to give 85 mg (50%) of
thin red needles. MS m/z (relative intensity): 535.05 (100%), 534.10
(97.5%), 533.10 (82.2%), 536.10 (46.4%), 537.05 (42.5%). 1H NMR
(400 MHz, CDCl3, δ): 10.33 (dd, J ) 6.4, 1.6 Hz, 2H), 7.90 (td,
J ) 7.2, 2.0 Hz, 2H), 7.58 (dd, J ) 8.0, 1.6 Hz, 2H), 7.02 (td, J )
6.8, 1.6 Hz, 2H), 2.71 (q, J ) 7.6 Hz, 4H), 1.27 (t, J ) 7.6 Hz,
6H). 13C NMR (100 MHz, CDCl3, δ): 154.3, 152.4, 150.3, 145.3,
137.8, 127.8, 119.7, 18.0, 14.5. HRMS-FAB (m/z): [M + H]+ calcd
for C18H18N5PtCl, 534.0950; found, 534.0944. Anal. Calcd for
C18H18N5PtCl: C, 40.42; H, 3.39; N, 13.09. Found: C, 40.70; H,
3.40; N, 12.33.
1,3-Bis(2-pyridylimino)benz(f)isoindolate Platinum(II) Chlo-
ride (3). Amounts of 0.50 g (1.34 mmol) of (COD)PtCl2 and 0.433
g (1.24 mmol) of benz(f)BPI were suspended in 25 mL of
methanol. To this solution was added 0.186 mL (1.34 mmol) of
triethylamine, and the mixture was heated to 50 °C under nitrogen
for 24 h. Upon cooling the reaction mixture to room temperature,
precipitate began to form. The precipitate was collected by filtration
and washed with water to give 0.554 g (78%) of a pale orange
solid. Sample was further purified by sublimation (300 °C, ∼10-4
Torr). MS m/z (relative intensity): 579.05 (100%), 578.10 (97.4%),
1
577.10 (79.3%), 580.10 (48.3%), 581.05 (40.0%). H NMR (400
Experimental Section
(10) McDermott, J. X.; White, J. F.; Whitesides, G. M. J. Am. Chem. Soc.
1976, 98, 6521–6528.
General Information. 3-Hexyne, 2-aminopyridine, 1,2-dicy-
anobenzene, 1-aminoisoquinoline, bromine, copper(I) cyanide, 1,5-
(11) Pincock, J. A.; Yates, K. Can. J. Chem. 1970, 48, 3332–3348.
(12) Fitzgerald, J.; Taylor, W.; Owen, H. Synthesis 1991, 9, 686–688.
(13) Meder, M.; Galka, C. H.; Gade, L. H. Monatsh. Chem. 2005, 136,
1693–1706.
(8) Martin, N.; Segura, J. L.; Seoane, C. J. Mater. Chem. 1997, 7, 1661–
1676.
(9) Dreuw, A.; Head-Gordon, M. Chem. ReV. 2005, 105, 4009–4037, and
(14) Baird, D. M.; Maehlmann, W. P.; Bereman, R. D.; Singh, P. J. Coord.
Chem. 1997, 42, 107–126.
references therein.
9
16248 J. AM. CHEM. SOC. VOL. 132, NO. 45, 2010

Blue- or Red-Shifted Absorption of Small Molecules
A R T I C L E S
MHz, CDCl3, δ): 10.3 (dd, J ) 6.8, 1.6 Hz, 2H), 8.59 (s, 2H),
8.09 (dd, J ) 6.0, 3.6 Hz, 2H), 7.95 (td, J ) 7.2, 1.6 Hz, 2H),
7.69-7.62 (m, 4H), 7.05 (td, J ) 6.8, 1.6 Hz, 2H). HRMS-FAB
(m/z): [M + H]+ calcd for C22H14N5196PtCl, 580.0660; found,
580.0642. Anal Calcd for C22H14N5PtCl: C, 45.64; H, 2.44; N,
12.10. Found: C, 45.50; H, 2.31; N, 11.69.
1,3-Bis(2-pyridylimino)benz(e)isoindole (benz(e)BPI). A solu-
tion of 1.0 g (5.6 mmol) of 1,2-dicyanonaphthalene, 1.09 g (11.7
mmol) of 2-aminopyridine, and 0.124 g (1.12 mmol) of CaCl2 in
20 mL of 1-butanol was refluxed under N2. The reaction was
monitored with TLC for the disappearance of 1,2-dicyanonaphtha-
lene. After 20 days of refluxing, the reaction was discontinued even
though starting material was still observed. Upon cooling the
reaction mixture to room temperature, precipitate began to form.
The precipitate was collected by filtration and washed with water.
The product was then isolated by column chromatography on silica
gel eluting first with CH2Cl2, then slow addition of ethylacetate to
the eluting solvent. Three fractions were isolated: first was 1,2-
dicyanonaphthalene (Rf ) 0.6, CH2Cl2), second was 1-(2-py-
ridylimino) isoindol-3-amine (Rf ) 0.3, CH2Cl2), and finally a
yellow fraction containing 1,3-bis(2-pyridylimino)benz(e)isoindole
(Rf ) 0.1, CH2Cl2). The yellow fraction was rotary evaporated to
Figure 1. CV of complexes 2 and 5 (150 mV/s). The peaks at 0 V are due
to the internal ferrocene reference.
Calcd for C26H16N5PtCl: C, 49.65; H, 2.56; N, 11.13. Found: C,
49.59; H, 2.41; N, 10.72.
Electrochemical and Photophysical Characterization. All
cyclic voltammetry (CV) and differential pulse voltammetry (DPV)
were performed using a EG&G Potentiostat/Galvanostat model 283.
DMF (VWR), distilled from Type 4A 1/16” molecular sieves (Alfa
Aesar), was used as the solvent under inert atmosphere with 0.1 M
tetra(n-butyl)ammonium hexafluorophosphate (Aldrich) as the sup-
porting electrolyte. A glassy carbon rod, a platinum wire, and a
silver wire were used as the working electrode, the counter
electrode, and the pseudo reference electrode, respectively. Elec-
trochemical reversibility was determined using CV, while all redox
potentials were found using DPV and reported relative to a
ferrocenium/ferocene (Fc+/Fc) redox couple used as an internal
standard.15
The UV-visible spectra were recorded using an Agilent 8453
UV-visible photo diode array spectrophotometer. Steady-state
emission experiments were performed on a Photon Technology
International QuantaMaster model C-60 spectrofluorimeter. All
lifetime measurements were performed on a IBH Fluorocube
lifetime instrument by a time-correlated single-photon counting
method using a 405-nm LED excitation source. Quantum efficiency
measurements were carried out using a Hamamatsu C9920 system
equipped with a xenon lamp, calibrated integrating sphere, and
model C10027 photonic multichannel analyzer.
Computational Methods. All properties reported here were
determined with first principles electronic structure calculations
using the GAMESS electronic structure code.16 To clearly dem-
onstrate a possible origin of the observed hypsochromic behavior,
properties are predicted and analyzed for 2,5-bis(2-pyridylimi-
no)pyrrolate platinum(II) chloride (1′) rather than the diethyl
substituted 2,5-bis(2-pyridylimino)3,4-diethylpyrrolate platinum(II)
chloride structure (1). Geometry optimizations for 1′ and 2-5 (see
Figure 1) were calculated with density functional theory (DFT),
employing the hybrid B3LYP functional.17 Platinum and chlorine
are described using small-core model core potentials MCPtzp
(triple-ꢀ + polarization basis set) and MCPdzp (double-ꢀ +
1
dryness to give 0.320 g (15%) of a yellow solid. H NMR (250
MHz, CDCl3, δ): 9.65 (d, J ) 8.25, 1H), 8.64 (s, 2H), 8.18-8.04
(m, 2H), 7.97 (d, J ) 8.25 Hz, 1H), 7.89-7.45 (m, 6H), 7.20-7.08
(m, 2H).
1,3-Bis(2-pyridylimino)benz(e)isoindolate Platinum(II) Chlo-
ride (4). Amounts of 0.1 g (0.268 mmol) of (COD)PtCl2 and 0.085
g (0.243 mmol) of benz(e)BPI were suspended in 10 mL of
methanol. To this solution was added 0.037 mL (0.268 mmol) of
triethylamine, and the mixture was heated to 50 °C under nitrogen
for 24 h. Upon cooling to room temperature, precipitate began to
form. The precipitate was collected by filtration and washed with
methanol. The red solid was then dissolved in boiling toluene and
then cooled to -40 °C overnight. The red powder was then collected
by filtration and washed with MeOH to give 0.062 g (44%) of the
desired product. MS m/z (relative intensity): 579.05 (100%), 578.10
(99.8%), 577.10 (82.5%), 580.10 (49.3%), 581.05 (41.9%). 1H NMR
(400 MHz, CDCl3, δ): 10.37 (s, 2H), 9.67 (d, J ) 8 Hz, 1H), 8.17
(d, J ) 8.4 Hz, 1H), 8.09 (d, J ) 8.4 Hz, 1H), 7.86-7.60 (m, 7H),
7.08 (d, J ) 6.8 Hz, 2H). HRMS-FAB (m/z): [M + H]+ calcd for
C22H14N5194PtCl, 578.0637; found, 578.0627. Anal. Calcd for
C22H14N5PtCl: C, 45.64; H, 2.44; N, 12.10. Found: C, 45.43; H,
2.18; N, 11.67.
1,3-Bis(1-isoquinolylimino)isoindole (BIQI). A mixture of
0.421 g (3.29 mmol) of 1,2-dicyanobenzene, 1 g (6.9 mmol) of 1-
aminoisoquinoline, and 0.11 g (1 mmol) of CaCl2 in 20 mL of
1-butanol was refluxed under N2 for 5 days. Upon cooling the
reaction mixture to room temperature, precipitate began to form.
The precipitate was collected by filtration and washed with water
to give 1.091 g (83%) of a green solid. The sample was used without
further purification for the next reaction. 1H NMR (250 MHz,
CDCl3, δ): 9.03 (d, J ) 8 Hz, 2H), 8.57 (d, J ) 5.75 Hz, 2H), 8.28
(dd, J ) 5.5, 3 Hz, 2H), 7.88-7.63 (m, 8H), 7.53 (d, J ) 5.75 Hz,
2H).
1,3-Bis(1-isoquinolylimino)isoindolate Platinum(II) Chloride
(5). Amounts of 0.360 g (0.965 mmol) of (COD)PtCl2 and 0.356 g
(0.893 mmol) of BIQI were suspended in 20 mL of methanol. To
this solution was added 0.134 mL (0.965 mmol) of triethylamine,
and the mixture was heated to 50 °C under nitrogen for 24 h. Upon
cooling the reaction mixture to room temperature, precipitate began
to form. The precipitate was collected by filtration and washed with
water to give 0.505 g (90%) of a dark purple solid. Sample was
further purified by sublimation (350 °C, ∼10-4 Torr). Because of
the low solubility of the product, NMR data were not obtained.
MS m/z (relative intensity): 628.10 (100%), 629.05 (82.4%), 627.00
(61.2%), 630.00 (47.7%), 631.10 (37.9%). HRMS-FAB (m/z): [M
+ H]+ calcd for C26H16N5PtCl, 628.0794; found, 628.0792. Anal.
(15) Gagne, R. R.; Koval, C. A.; Lisensky, G. C. Inorg. Chem. 1980, 19,
2854–2855.
(16) (a) Schmidt, M. W.; Baldridge, K. K.; Boatz, J. A.; Elbert, S. T.;
Gordon, M. S.; Jensen, J. H.; Koseki, S.; Matsunaga, N.; Nguyen,
K. A.; Su, S.; Windus, T. L.; Dupuis, M.; Montgomery, J. A. J.
J. Comput. Chem. 1993, 14, 1347–1363. (b) Gordon, M. S.; Schmidt,
M. W. Theories and Applications of Computational Chemistry, the
First Forty Years; Elsevier: Amsterdam, 2005.
(17) (a) Hohenberg, P. Phys. ReV. 1964, 136, B864–871. (b) Kohn, W.;
Sham, L. J. Phys. ReV. 1965, 140, A1133–1138. (c) Becke, A. D.
J. Chem. Phys. 1993, 98, 5648–5652. (d) Hertwig, R. H.; Koch, W.
Chem. Phys. Lett. 1997, 268, 345–351. (e) Stephens, P. J.; Devlin,
F. J.; Chabalowski, C. F.; Frisch, M. J. J. Phys. Chem. 1994, 98,
11623–11627.
9
J. AM. CHEM. SOC. VOL. 132, NO. 45, 2010 16249

A R T I C L E S
Hanson et al.
polarization basis set), respectively,18 whereas all remaining atoms
are treated with the all-electron cc-pVDZ basis set.19 All stationary
points were confirmed as minima on the ground-state potential
energy surface by calculating and diagonalizing the Hessian (matrix
of energy second derivatives). Vertical excitation energies were
calculated with time-dependent DFT (TDDFT) using the B3LYP
functional. The effects of solvent on the geometry and vertical
excitation energies were studied using the Conductor-like Polariz-
able Continuum Model (CPCM).20
Table 1. Electrochemical Potentials for 1-5 Reported in Volts (V)
Relative to Fc+/Fca
b
red2
red1
complex
E1/2
E1/2
Eox
∆E1/2
1
2
3
-1.89
-1.87
-1.95
-1.73
-1.58
-1.41
-1.44
-1.53
-1.32
-1.23
0.79
0.79
0.80
0.77
0.79
2.20
2.22
2.33
2.09
2.07
4
5c
The Tamm-Dancoff approximation (TDA)21 to TDDFT is used
here for the purposes of conveniently analyzing the TDDFT
excitation energies and validating an orbital analysis. As noted by
Dreuw and Head-Gordon,9 application of the TDA to the DFT linear
response equations (TDA-TDDFT) provides an analogue to the
simple configuration interaction with single excitations (CIS)
method. This is a crucial feature, because an excited state dominated
by a single CIS configuration can be approximately characterized
using simple orbital arguments. Similar use of an orbital argument
is applied here if the TDA-TDDFT method predicts an excited state
dominated by a single excitation (e.g., HOMOfLUMO).
a Measurements were performed in an anhydrous solution of 0.1 M
NBu4PF6 in DMF. b ∆E1/2 ) Eox - E1/2
.
c Two additional reversible
red1
reduction waves were observed at -1.62 and -1.93 V.
complexes 2 and 5 are shown in Figure 1. All five complexes
display irreversible oxidation waves at ∼0.8 V. A small return
wave near 0.80 V can also be observed for complexes 1-4;
however, full reversibility was not apparent at any of the tested
scan rates (50-500 mV/s).
Multiple reversible reduction waves were observed in all five
complexes. The complexes with pyridyl groups (1-4) display
two reversible waves. For complexes 1-3, the first reduction
peak shifts to more negative values with each successive
benzannulation. The trend is counter to the common expectation
that the species with the larger π-orbital system will provide a
greater stabilization to the negative charge.23 It is worth noting
that the addition of butadiene at the 6,7- and the 5,6-positions
of 2 results in a negative shift (3; -1.53 V) and a positive shift
(4; -1.32 V) in reduction potential relative to 2 (-1.44 V).
For complex 5, a total of four reversible reduction peaks were
Results and Discussion
Synthesis. The ligand precursors for complexes 2-5 (BPI,
benz(f)BPI, benz(e)BPI, and BIQI, respectively) were synthe-
sized in 15-83% yield using a procedure, metal-ion-catalyzed
nucleophilic addition of aromatic amines to dicyano aromatics,
developed by Siegl.7 The precursor for complex 1 (BPEP) was
synthesized by bromination of 3-hexyne,11 followed by substitu-
tion with cyanide, photoisomerization,12 and the nucleophilic
addition procedure mentioned above. Photoisomerization of (E)-
3,4-dicyano-3-hexene was required because all attempts to
produce BPEP from the (E)-isomer were unsuccessful. The
significantly lower yield of BPEP, 4%, may be due to thermal
isomerization from (Z)-3,4-dicyano-3-hexene to the unreactive
(E)-isomer.12
The complexes 1-5 were synthesized in 44-90% yield using
(COD)PtCl2 as a metal precursor, rather than the more com-
monly used (PhCN)2PtCl2, in accord with the results found by
Meder et al.13 The complexes were characterized by mass
spectrometry, elemental analysis, and NMR spectroscopy.
Complexes 1-5 are soluble in common organic solvents,
although increased benzannulation decreases solubility. Differ-
red
observed (E1/2 ) -1.23, -1.58, -1.62, -1.93 V). The two
new reduction waves are likely due to additional reduction sites
on the isoquinolyl moieties. Like 4, complex 5 displays a
positive shift in reduction potential relative to 2, as opposed to
the negative shift observed for 3. This shift in potential leads
to a decrease in ∆E1/2 for 5 and 4, relative to the trend of
increasing ∆E1/2 observed in the series 1-3 (Table 1).
Electronic Spectroscopy. The absorption spectra for com-
plexes 1-5 and their ligand precursors were recorded in CH2Cl2
and are shown in Figure 2. The BPI ligand (Figure 2a) displays
several absorption bands between 300-425 nm that are assigned
to π-π* transitions. The spectra for the other four ligands
display π-π* transitions with a similar vibrational progression
(Figure 2a). Upon platination of BPI, the ligand-centered (LC)
π-π* transitions (ε ≈ 2 × 104 M-1 cm-1) undergo a red shift,
while a new, low energy transition appears between 425-550
nm (ε ≈ 1.4 × 104 M-1 cm-1). The low energy absorption band
is assigned to a combined metal-to-ligand (ML) and intraligand
(IL) charge transfer (CT) transition in accord with the prior
report by Wen et al.6 Similarly, distinct LC and ML-ILCT
absorption transitions are also observed in the spectra of 1 and
3-5 (Figure 2b).
Both the absorption onset and the overall peak position of
the ML-ILCT transition in 1-3 display a clear trend: a blue
shift with each successive benzannulation of the pyrrolate group.
This trend is in agreement with the observations of increasing
∆E1/2 in the series, but is counter to the common expectation
that expansion of the aromatic π-system will lead to red-shifted
absorption.3 It is also noteworthy that benzannulation of the
isoindole ring of 2 results in either a blue shift (in 3) or a red shift
(in 4) depending on the site of attachment. These observed trends
in absorption are in agreement with the change in ∆E1/2 values.
ences in solubility can most clearly be observed by the quality
1
of the H NMR spectra. Well-resolved 195Pt satellites (JPt-H
)
20.5 Hz) for the pyridyl hydrogens in the meta-position, similar
to those reported for [(2,2′;6′,2′′-terpyridine)PtCl]+,22 can be
identified in the highly soluble, diethyl substituted complex 1
(Figure S1). However, due to the low solubility of 5, we were
1
unable to obtain an acceptable H NMR spectrum for this
complex. Therefore, the identity and purity of 5 was confirmed
by mass spectrometry and elemental analysis.
Electrochemistry. The electrochemical properties of the
complexes 1-5 were examined using cyclic voltammetry (CV)
and differential pulsed voltammetry (DPV). The results of these
measurements are listed in Table 1, representative scans for
(18) Mori, H.; Ueno-Noto, K.; Osanai, Y.; Noro, T.; Fujiwara, M.;
Klobukowski, M.; Miyoshi, E. Chem. Phys. Lett. 2009, 476, 317–
322.
(19) Dunning, T. H., Jr. J. Chem. Phys. 1989, 90, 1007–1023.
(20) Cossi, M.; Barone, V. J. Chem. Phys. 2001, 115, 4708–4717.
(21) (a) Fetter, A. L.; Walecka, J. D. Quantum Theory of Many-Particle
Systems; McGraw-Hill: New York, 1971. (b) Hirata, S.; Head-Gordon,
M. Chem. Phys. Lett. 1999, 314, 291–299. (c) Tamm, I. J. Phys.
(USSR) 1945, 9, 449. (d) Dancoff, S. M. Phys. ReV. 1950, 78, 382–
385.
(23) Brooks, J.; Babayan, Y.; Lamansky, S.; Djurovich, P. I.; Tsyba, I.;
(22) Cummings, S. D. Coord. Chem. ReV. 2009, 253, 449–478.
Bau, R.; Thompson, M. E. Inorg. Chem. 2002, 41, 3055–3066.
9
16250 J. AM. CHEM. SOC. VOL. 132, NO. 45, 2010

Blue- or Red-Shifted Absorption of Small Molecules
A R T I C L E S
of 1-5 parallel the shifts observed in the absorption spectra
that occur with each successive benzannulation. Interestingly,
the complexes that show the most red-shifted emission (λmax
650 nm) have both the smallest (1) and the largest (5) π-systems.
≈
Computational Results. The geometry optimized structure of
2 (Figure 4) is in good agreement with the X-ray crystal structure
obtained by Meder et al. for the alkyl substituted analogue (6-
Me-3-tBuBPI)PtCl (Table 3).13 The calculated metal-ligand
bond lengths differ by less than 0.05 Å, while the chelate bond
angles differ by less than 0.5°. The calculations accurately
describe an out of the plane distortion (∼10°) of the chlorine
atom caused by steric conflict with the ortho-hydrogens (at
position 1) of the pyridyl rings (Cl · · ·H ) 2.24 Å). Similar bond
lengths and angles in the coordination sphere were found in
the calculated structures of complexes 1′ and 3-5 (Table S1).
From the optimized structures, vertical excitation energies
are determined using time-dependent density functional theory
(TDDFT). The lowest energy vertical transition for both gas
phase and solution (CPCM) calculations (Table 4) show
qualitative agreement with the experimental absorption trends
(Table 2). The CPCM calculations are in better quantitative
agreement with the experimental data than are the gas-phase
predictions. As is observed in the experiments for complexes
1-3, a blue shift in absorption is predicted with each successive
benzannulation at the pyrrole position. Although complexes 4
and 5 are also benzannulated versions of 2, they display a red
shift in absorption relative to 2, as opposed to the blue shift
observed for 3. Also listed in Table 4 are the contributions of
the HOMO-LUMO configurations to the TDDFT excitation
energies. Because the sum of the squares of all such coefficients
is unity, it is clear that the HOMOfLUMO contribution
dominates these transitions (82-92%). Note also that the TDA-
TDDFT vertical excitation energies are in excellent agreement
with the TDDFT predictions. This is important: Because the
TDA-TDDFT method is essentially equivalent to the CIS wave
function approach, the large coefficients in Table 4 provide
credence to the use of a HOMO-LUMO argument for
qualitatively understanding the observed and predicted trends.24
Because the HOMOfLUMO transition dominates the excita-
tions that are of interest here, interpretations of the predictions
and observations are likely to employ these frontier orbitals.
The Kohn-Sham HOMO and LUMO energies for the
ground-state geometries are given in Table 5. The trends in
Table 5 correspond closely with those observed in the electro-
chemistry experiments (Table 1). The calculated HOMO energy
levels vary by no more than 0.03 eV, qualitatively agreeing with
the near equivalent oxidation potentials for 1-5. The HOMOs
of 1′ and 2-5 are predominantly localized on the [(py-
ridylimino)3,4-pyrrolate]PtCl portion of the molecules (Figures
5 and 6), and thus the HOMO energies appear to be dictated
by this moiety. In contrast to the invariance in HOMO energies,
the energy of the LUMO increases in going from 1′ to 2 to 3.
This trend corresponds to a more negative reduction potential
with each successive benzannulation, as is documented in Table
1. On the other hand, the LUMO energies decrease from 2 to
4 to 5 and reflect the values observed for the reduction potentials
of each respective complex.
Figure 2. (a) Normalized absorption spectra of free ligands: BPEP, BPI,
benz(f)BPI, benz(e)BPI, and BIQI. The metal complex for each ligand is
given in parentheses. (b) Absorption spectra of 1-5 at room temperature
in CH2Cl2.
Figure 3. Emission spectra of 1-5 in 2-MeTHF at 77 K.
All of the complexes were luminescent in glassy solvent (2-
MeTHF) at 77 K (Figure 3), and complexes 2-5 were also
emissive at room temperature. Emission data for 1-5 are
summarized in Table 2. The spectra recorded at 77 K display
vibronic structure and exhibit long lifetimes (τ ) 0.081-20.8
µs), arising from
a
mixture of 3[πfπ*(L)] 3IL and
3[5d(Pt)fπ*(L)] MLCT triplet excited states as assigned by
Wen et al.6 The quantum yields at room temperature for
complexes 2-5 varied between 0.3% and 11.5%, whereas 1
was nonemissive and only weakly luminescent even at 77 K.
The emission properties of 2 and 3 are in accord with data
reported by Wen et al.,6 although we observe a somewhat higher
value for the quantum yield of 2 (Φ ) 0.015 vs 0.005) and 3
(Φ ) 0.1 vs 0.04) in CH2Cl2 (Table 2) presumably due to
differences in measurement techniques. The emission energies
3
Both the theoretical and the experimental results suggest that
the uncharacteristic blue shift in absorption going from 1 to 2
to 3 is due to the destabilization of the LUMO with successive
(24) Of course, unoccupied orbitals are not variationally determined, so
arguments based on, for example, a LUMO are only qualitative.
9
J. AM. CHEM. SOC. VOL. 132, NO. 45, 2010 16251

A R T I C L E S
Hanson et al.
Table 2. Photophysical Properties of Complexes 1-5
emission at rtb
emission at 77 Kc
kr
knr
(105 s-1
a
)
d
)
e
)
complex
absorbance λ (nm) (ε, ×104 M-1 cm-1
λmax (nm)
τ (µs)
ΦPL
(104 s-1
λmax (nm) τ (µs)
1
2
3
4
5
234(2.32), 278(2.14), 385(1.26), 500(0.73)
650
595
566
638
648
0.081
6.8
20.8
1.04
2.0
250(5.37), 277(3.51), 348(2.24), 473(1.36), 487(1.35)
229(4.16), 300(5.03), 387(1.78), 409(1.89), 467(1.20), 477(1.21)
243(2.93), 280(2.95), 376(1.40), 402(1.37), 502(0.57), 529(0.46)
245(5.51), 262(5.48), 382(1.61), 403(1.55), 426(1.29), 511(1.75), 536
(1.62)
641
611
682
711
0.99 (0.97) 0.014 (0.015)
1.4
2.6
2.0
5.9
9.96
1.97
66.5
195.5
4.5 (3.2)
0.15
0.115 (0.098)
0.003
0.051
0.003
e
a In CH2Cl2. b In toluene deaerated with N2. Data in parentheses recorded in CH2Cl2. c In 2-MeTHF. d kr ) Φ/τ. knr ) (1 - Φ)/τ.
Table 5. HOMO/LUMO Energies and the Kohn-Sham Shifts
(HOMO-LUMO Gap) for the Gas Phase and CPCM Solvated
Complexes
HOMO
energy (eV)
LUMO
energy (eV)
Kohn-Sham orbital
energy shift (eV)
complex
1′
-5.55
-5.55
-5.52
-5.52
-5.52
-5.74
-5.77
-5.77
-5.74
-5.74
-2.78
-2.56
-2.45
-2.69
-2.78
-2.83
-2.64
-2.56
-2.78
-2.86
2.78
2.99
3.07
2.83
2.75
2.91
3.13
3.21
2.96
2.88
2
3
4
5
CPCM-1′
CPCM-2
CPCM-3
CPCM-4
CPCM-5
Figure 4. Calculated structure of 2, with atoms carbon and nitrogen
indicated by black and blue spheres respectively. Hydrogen atoms were
omitted for clarity.
a more quantitative analysis of the LUMO destabilization is
presented, using simple molecular orbital theory that relies on
both energetic and symmetry considerations of the benzannu-
lation process similar to that used by Uno et al. to explain the
observed more negative reduction potentials of p-quinones
caused by benzannulation.25
Table 3. Selected Bond Lengths (Å) and Angles (deg) from DFT
Calculations and Reported X-ray Data of 2
B3LYP
CPCM/B3LYP
expa
Pt-Cl (Å)
2.36
2.09
2.00
2.37
2.09
2.00
2.335(2)
2.049(4)
1.966(3)
Pt-Npyr (Å)
Pt-Nind (Å)
Cl-Pt-Nind (deg)
One can visualize the formation of the frontier molecular
orbitals of 2 by combining the valence orbitals of 1′ with those
of 1,3-butadiene as illustrated in Figure 5. The HOMO of 1′
consists of contributions from the respective p- and d-orbitals
of the chloride and platinum atoms, as well as from the π-system
of the pyridyl and imino-pyrrolate portions of the ligand. There
is minimal HOMO density at the site of benzannulation of 1′
by 1,3-butadiene; thus no orbital mixing is observed and the
HOMO energy remains unchanged in 2. The LUMO of 1′ is
localized primarily on the π-system of the ligand with only a
small contribution from the d-orbital on the Pt atom. If one
ignores the out of plane distortion of the chloride atom, 1′ can
be idealized as having C2V symmetry. In this point group, the
LUMO of 1′ can be considered to have a2 symmetry. The
frontier orbitals of cis-1,3-butadiene also fall under the C2V point
group with the HOMO and LUMO designated as a2 and b1,
respectively. The LUMO of 1,3-butadiene (b1) does not have
the appropriate symmetry to mix with the LUMO of 1′ (a2).
However, both the HOMO of 1,3-butadiene and the LUMO of
1′ have the same a2 symmetry, so they can combine to create
an occupied bonding MO (not shown) and an unoccupied
antibonding orbital (LUMO of 2) with the addition of a new
nodal plane at the site of attachment. The favorable orbital
symmetry enables the HOMO of 1,3-butadiene to act as an
effective electron-donating group to the LUMO of 1′. The net
effect of these interactions is an unaltered HOMO and a
destabilized LUMO, resulting in the blue-shifted absorption
upon benzannulation of 1′. Likewise, a similar combination
between the frontier orbitals of 2 and 1,3-butadiene leads to an
171.6
170.5
170.14(12)
a From ref 13 (pyr ) pyridine, ind ) indolate).
Table 4. TDDFT and TDA Vertical Excitation Energies for Gas
Phase and CPCM Solvated Complexes
TDDFT vertical
excitation energy
nm (error)a
TDDFT TDA
vertical excitation
energy nm (error)a
square of
HOMOfLUMO
TDDFT coefficients
complex
1′
541 (-)
519 (32)
498 (21)
564 (35)
575 (39)
541 (-)
489 (2)
472 (5)
533 (4)
540 (4)
566 (-)
508 (21)
488 (11)
553 (24)
564 (28)
525 (-)
478 (9)
462 (15)
522 (7)
527 (11)
0.83
0.85
0.85
0.87
0.83
0.82
0.84
0.86
0.92
0.82
2
3
4
5
CPCM-1′
CPCM-2
CPCM-3
CPCM-4
CPCM-5
a The absolute difference between calculated and experimental
transitions, reported in nanometers, is given in parentheses.
expansion of the π-system of the pyrrolate moieties. In a recent
report, the destabilization of the LUMO of 3 relative to 2 was
rationalized using energetic considerations and bonding/anti-
bonding interactions between the (2-pyridylimino)3,4-pyrrolate
and naphthyl portions of the molecules. However, we find that
combining a naphthyl fragment with a (2-pyridylimino)3,4-
pyrrolate moiety leads to either a destabilization (in 3) or a
stabilization (in 4) of the LUMO for the two isomers. Clearly
the energy of the LUMO in 3 and 4 is not dictated solely by
the LUMO energy of naphthalene, and thus the model proffered
by Wen et al. does not explain the observed phenomena.6 Here,
(25) Uno, B.; Kano, K.; Konse, T.; Kubota, T.; Matsuzaki, S.; Kuboyama,
A. Chem. Pharm. Bull. 1985, 33, 5155–5166.
9
16252 J. AM. CHEM. SOC. VOL. 132, NO. 45, 2010

Blue- or Red-Shifted Absorption of Small Molecules
A R T I C L E S
Figure 5. Qualitative orbital diagram of the valence orbitals for complexes 1′, 2, and 3. The HOMO (bottom, solid) and LUMO (top, transparent) surfaces
are displayed as viewed above the π-symmetric orbitals, with opposite phases above and below the plane of the molecule.
Figure 6. Qualitative orbital diagram of the valence orbitals for complexes 2, 4, and 5.
unaltered HOMO and a destabilized LUMO as seen in 3 (Figure
5). Even when the geometry distorts somewhat from the
idealized C2V symmetry, the nodal behavior of the orbitals at
the site of butadiene addition can be used to make the same
qualitative arguments. This is expanded upon in the following
paragraphs.
The arguments regarding the symmetry of interacting orbitals
can also be used to interpret the effects of benzannulation at
other positions of 2. The stabilizing effect on the LUMO when
2 is benzannulated at either the 5,6- or the 3,4-positions (to
form 4 and 5, respectively) is illustrated in Figure 6. The orbitals
of 2 can be characterized by the presence or absence of a
bisecting nodal plane at the site of benzannulation. The
absence of a perpendicular bisecting nodal plane at either of
the relevant positions of the LUMO of 2 favors a cooperative
interaction with the LUMO of 1,3-butadiene that leads to a
bonding/antibonding pair of MOs, whereas interactions with
the HOMO of 1,3-butadiene are disfavored. The end result
is a stabilized LUMO in both 4 and 5. Analogous symmetry
considerations predict that benzannulation at either the 1,2-
or the 2,3- position of 2 should stabilize and destabilize the
LUMO, respectively, an outcome that is supported by
calculation (Figure S7).
The unexpected destabilization of the LUMO, and corre-
sponding increase in reduction potential, exhibited by 2 and 3
has been observed in other molecules that have their π-systems
extended with either 1,3-butadiene or ethene. These examples
include tetracyanoquinones (TCNQ),8 indoanilines,5c anhy-
drides,5d Ru(polypyridyl)3,5a complexes and phthalocyanines,26
as well as in simple polycyclic aromatic hydrocarbons.27 The
destabilization of the LUMO has been attributed to either
molecular distortions (in TCNQ),8 an increase in the total
9
J. AM. CHEM. SOC. VOL. 132, NO. 45, 2010 16253

A R T I C L E S
Hanson et al.
Chart 2
LUMO destabilization is small relative to the HOMO destabi-
lization; thus an overall red shift is observed. Similarly, DFT
calculations by Nguyen and Pachter show that the LUMOs of
tetra-azaporphyrins are destabilized when benzannulated to form
phthalocyanines.26
Thus far, the orbital symmetry analysis used to explain the
destabilization of the LUMO has been limited to benzannulation
with 1,3-butadiene; however, the same principles can also be
applied to molecules whose π-systems are expanded by the
addition of ethene. For example, Adachi, et al. found that
benzannulation with ethene at the 1,12- and 6,7-positions of
3,4,9,10-perylenetetracarboxylic dianhydride diimide (PTCAI,
Chart 2) to give coronenetetracarboxylic dianhydride diimide
(CTCAI) resulted in an increase in the LUMO energy.5d The
LUMO of PTCAI (b2g symmetry) has an in-phase pair of orbitals
at the 1,12- and 6,7-positions that can combine with the
π-symmetric HOMO, but not the π-antisymmetric LUMO, of
ethene in a fashion similar to that shown in Figure 5 for 2 (see
Figure S8). The favorable interaction between the LUMO of
PTCAI and the HOMO of ethene will consequently destabilize
the LUMO of PTCAI. The upward shift in the LUMO energy
that occurs upon ethene addition to PTCAI has a parallel in the
change in the reduction potentials found for perylene (E1/2
)
antibonding character of the orbital (in indoanilines and
anhydrides),5c,d or an increased localization of the π* orbital
(in Ru(polypyridyl) complexes).5a The molecular orbital model
presented here can be successfully applied to all of these
molecular systems to rationalize the origin of the unusual LUMO
destabilization by the same sort of orbital mixing approach
described for compounds 2 and 3.
-1.64 V vs SCE) versus corannulene (E1/2 ) -2.03 V vs
SCE).27
A slightly more complex situation alters the HOMO of
PTCAI when the π-system is expanded to form CTCAI. The
HOMO of PTCAI is stabilized by a bonding/antibonding
interaction between the π-antisymmetric LUMO of ethene and
the out-of-phase pair of orbitals at the 1,12- and 6,7-positions
in the HOMO (au) of PTCAI (Figure S8). On the other hand,
the HOMO-4 (b1u) of PTCAI has an in-phase pair of orbitals
at the same positions, and, as a result, this orbital is strongly
destabilized by an antibonding interaction with the π-symmetric
HOMO of ethene. The energy of the HOMO-4 of PTCAI is
so destabilized that it becomes the HOMO of CTCAI. The result
of these interactions is that the HOMO (LUMO) of CTCAI is
stabilized (destabilized) relative to the parent PTCAI. Thus, a
net blue shift in the lowest energy transition from 526 nm in
PTCAI28 to 511 nm in CTCAI is observed.29
The arguments given above to explain the stabilization/
destabilization of the LUMO energies upon benzannulation can
also be used to rationalize related shifts in HOMO energies in
molecules that serve as electron donors. For example, tetrathi-
afulvalene (TTF, E1/2 ) 0.35 V vs SCE) becomes more difficult
to oxidize upon benzannulation to form dibenzotetrathiafulvalene
(E1/2 ) 0.60 V vs SCE).5b Benzannulation of dibenzotetrathi-
afulvalene results in either an increase (1,2-position; E1/2 ) 0.72
V vs SCE)5b or a decrease (2,3-position; E1/2 ) 0.52 V vs
SCE)30 in oxidation potential depending on the orbital symmetry
at the site of butadiene addition (Figure 7). The atypical HOMO
stabilization upon benzannulation again results in a blue-shifted
absorption from 516 nm in dibenzotetrathiafulvalene to 502 nm
in the linear dinaphthotetrathiafulvalene.
As with complexes 1-3, benzannulation in several of these
systems leads to destabilization of the LUMO, which manifests
itself in both increased reduction potentials and blue-shifts in
the lowest energy absorption transitions. A necessary require-
ment to observe the blue-shift is for the HOMO energy to be
either stabilized or remain relatively unchanged upon benzan-
nulation. This condition is met when the HOMO is electronically
isolated from the site of benzannulation. For example, indoa-
niline dyes (Chart 2) have a HOMO localized on the p-diamino-
phenyl fragment and a LUMO localized on the p-quinoaniline
moiety.5c The low energy charge transfer absorption bands in
these molecules undergo a progressive shift from 596 nm to
589 nm to 558 nm when the 1,2- and 3,4-positions of the
quinoaniline fragment are successively benzannulated. The shift
to higher energy is attributed to an increase in LUMO energy
that accompanies each benzannulation, while the HOMO energy
remains relatively unchanged. Likewise, benzannulation at the
1,2-positions of (Ru(2,2′-bipyridine)3 leads to a 60 nm blue shift
in the lowest energy absorption band.5a This shift is the result
of LUMO destabilization as the reduction potential is shifted
by 0.16 V (E1/2 ) -1.51 V vs SCE in Ru(2,2′-biisoquinoline)3
vs -1.35 V in Ru(bpy)3). While the HOMO energy is stabilized
by 0.14 V, the change relative to the LUMO is smaller, and
thus a blue shift is observed.
Conclusion
Although the discussion so far has focused on molecules that
exhibit a blue-shifted absorption upon benzannulation, there are
several examples where LUMO destabilization occurs and red-
shifted absorption is observed. For example, theoretical inves-
tigations of porphyrins have shown that these molecules exhibit
a destabilized LUMO upon benzanulation.3b,26 The amount of
A series of (BPI)PtCl derivatives with varying degrees of
benzannulation have been prepared, and their properties have
(28) Adachi, M.; Murata, Y.; Nakamura, S. J. Phys. Chem. 1995, 99,
14240–14246.
(29) Rohr, U.; Schlichting, P.; Bohm, A.; Gross, M.; Meerholz, K.;
Brauchle, C.; Mullen, K. Angew. Chem., Int. Ed. 1998, 37, 1434–
1437.
(26) Nguyen, K. A.; Pachter, R. J. Chem. Phys. 2001, 114, 10757–10767.
(27) Jensen, B. S.; Parker, V. D. J. Am. Chem. Soc. 1975, 97, 5211.
(30) Schukat, G.; Fanghanel, E. J. Prakt. Chem. 1979, 321, 675–679.
9
16254 J. AM. CHEM. SOC. VOL. 132, NO. 45, 2010

Blue- or Red-Shifted Absorption of Small Molecules
A R T I C L E S
Figure 7. Qualitative orbital diagram for benzannulated dibenzotetrathiafulvalene complexes.
been characterized. It is observed that the lowest energy
absorption/emission transitions undergo a red or blue shift
depending on the site of benzannulation. Experimental data and
first principles calculations suggest that the observed trends are
caused by an unvaried HOMO energy level and a destabilized
or stabilized LUMO with benzannulation. A rationale for the
stabilization/destabilization of the LUMO is presented using
simple molecular orbital theory. The molecular orbital model
presented here can also be successfully applied to indoaniline
dyes, (Ru(2,2′-bipyridine)3)2+, 3,4,9,10-perylenetetracarboxylic
dianhydride diimide, and dibenzotetrathiafulvalene, to explain
unusual blue shifts in absorption with benzannulation.
Although the focus in this Article has been to describe blue shifts
in absorption upon benzannulation, the ability to selectively tune
HOMO and LUMO energy levels while extending conjugation may
be particularly useful in organic electronics. This concept is
exemplified in TTF systems mentioned above where the increased
field-effect transistor performance of dinaphthotetrathiafulvalene
over dibenzotetrathiafulvalene has been attributed to the more
extended π-system and its effects on film morphology.5b This
orbital analysis can also readily be applied to design new solar
cell and charge transfer materials where molecular organization
and the ability to tune the energy of the frontier orbitals of donor
and acceptor molecules is of the utmost importance.
based argument to explain the observed electronic properties
of a given molecule or series of molecules. The method
presented here relies on time-dependent DFT and related
theoretical methods and their correlation with experimental data.
In particular, this approach is based on the Tamm-Dancoff
approximation (TDA) to TDDFT. The orbitals predominantly
involved in this transition are determined from the coefficients
of the corresponding configurations for the lowest energy
excitation. If the TDA-TDDFT transition matches that derived
by TDDFT, and if the energy of the transition correlates with
experimental absorption energies, the application of an orbital
argument is valid. It was demonstrated that the molecules
investigated in the current work satisfy these conditions and
therefore warrant the use of molecular orbital interpretations.
Acknowledgment. K.H., P.I.D., and M.E.T. thank Universal
Display Corp. and the Department of Energy for their financial
support of this work. L.R., F.Z., and M.S.G. are grateful for the
support of the Air Force Office of Scientific Research.
1
Supporting Information Available: H NMR of 1-4, room
temperature emission spectra of 2-5, the orbital diagram for
PTCAI, and the products of benzanulation at the 2,3-position
and the 1,2-position of 2. This material is available free of charge
via the Internet at http://pubs.acs.org.
This article has also introduced a method that one can
routinely use to determine the validity of a molecular orbital-
JA1075162
9
J. AM. CHEM. SOC. VOL. 132, NO. 45, 2010 16255