Photodriven Electron and Energy Transfer from Copper Phenanthroline Excited States

Article abstract of DOI:10.1021/ic960503z

Electron and energy transfer from copper 1,10-phenanthroline excited states is observed at room temperature in organic solvents. The copper phenanthroline excited states are metal-to-ligand charge-transfer in nature and have lifetimes of ～70-250 ns in dic

Full text of DOI:10.1021/ic960503z

6406
Inorg. Chem. 1996, 35, 6406-6412
Photodriven Electron and Energy Transfer from Copper Phenanthroline Excited States
Mark Ruthkosky, Felix N. Castellano, and Gerald J. Meyer*
Department of Chemistry, Johns Hopkins University, Baltimore, Maryland 21218
ReceiVed May 8, 1996X
Electron and energy transfer from copper 1,10-phenanthroline excited states is observed at room temperature in
organic solvents. The copper phenanthroline excited states are metal-to-ligand charge-transfer in nature and have
lifetimes of ∼70-250 ns in dichloromethane solution if methyl or phenyl substituents are placed in the 2- and
9-positions of the phenanthroline ligand. The unsubstituted cuprous compound Cu(phen)2(PF6) is nonemissive
under these conditions, and the excited state lifetime is <20 ns. The rate and efficiency of energy transfer to
anthracene or electron transfer to viologens is reported. The cage escape efficiency of [Cu(dpp)22+, MV+•],
where dpp is 2,9-diphenyl-1,10-phenanthroline, is close to unity within experimental error. Back electron transfer
to ground state products occurs at the diffusion limit, 2 × 1010 M-1 s-1
.
Research into molecular solar energy conversion with inor-
ganic excited states has focused mainly on porphyrin and dπ6
transition metal chromophores.1 There has been growing
interest in the excited state and redox properties of cuprous
phenanthroline derivatives.2 These compounds possess broad
absorption bands in the visible region with high extinction
coefficients potentially useful for solar harvesting. Much of
the interest in the excited state properties of these compounds
was motivated by the pioneering work of McMillin and co-
workers, who first established that visible excitation of
Cu(dmp)2+ (dmp is 2,9-dimethyl-1,10-phenanthroline) produces
room-temperature photoluminescence in dichloromethane solu-
tion.3 This and related copper phenanthroline compounds with
alkyl or aryl substituents in the 2,9-positions of phenanthroline
also have long-lived excited states and display excellent
photostability.4-7 The excited states are potent reductants that
have been employed in potentially practical processes, such as
the production of hydrogen gas from water8 and an electrical
current by sensitization of wide band gap semiconductors.9
In a recent communication,10 we reported the first spectro-
scopic observation of photodriven energy transfer from excited
copper phenanthroline compounds. The acceptors were the
triplet states of anthracene, and the energy transfer dynamics
were quantitated by the appearance of the triplet-to-triplet
anthracene absorption, eq I. This process is important as it
results in a state with well-defined spin multiplicity, a 3-order-
of-magnitude increase in excited state lifetime, and little loss
in potential energy. In related processes, anthracene has been
employed as an energy transfer shuttle from ruthenium tris-
(bipyridine) (Ru(bpy)32+) to methyl viologen (MV2+) electron
acceptors.11 Here we explore related excited state electron and
energy transfer processes with a family of cuprous phenanthro-
line compounds.
X Abstract published in AdVance ACS Abstracts, October 1, 1996.
(1) Kalyanasundaram, K. Photochemistry of Polypyridine and Porphyrin
Complexes, Academic Press: London, 1992.
(2) For recent reviews of copper(I) excited states, see: (a) Kutal, C. Coord.
Chem. ReV. 1990, 99, 213. (b) Kalyanasundaram, K. Photochemistry
of Polypyridine and Porphyrin Complexes; Academic Press: London,
1992; Chapter 9. (c) Horvath, O. Coord. Chem. ReV. 1994, 135, 303.
(3) (a) McMillin, D. R.; Buckner, M. T.; Ahn, B. T. Inorg. Chem. 1977,
16, 943. (b) Blaskie, M. R.; McMillin, D. R. Inorg. Chem. 1980, 19,
3519. (c) Gamache, R. E.; Rader, R. A.; McMillin, D. R. J. Am. Chem.
Soc. 1985, 107, 1141. (d) Palmer, C. E. A.; McMillin, D. R.; Kirmaier,
C.; Holten, D. Inorg. Chem. 1987, 26, 3167. (e) Stacy, E. M.;
McMillin, D. R. Inorg. Chem. 1990, 29, 393. (f) Everly, R. M.;
McMillin, D. R. J. Phys. Chem. 1991, 95, 9071.
(4) (a) Kirchoff, J. R.; McMillin, D. R.; Marnot, P. A.; Sauvage, J. P. J.
Am. Chem. Soc. 1985, 107, 1138. (b) Dietrich-Buchecker, C. O.;
Marnot, P. A.; Sauvage, J. P.; Kintzinger, J. P.; Maltese, P. NouV. J.
Chim. 1984, 8, 573. (c) Ichinaga, A. K.; Kirchoff, J. R.; McMillin, D.
R.; Dietrich-Buchecker, C.O.; Marnot, P. A.; Sauvage, J. P. Inorg.
Chem. 1987, 26, 4290. (d) Gushurst, A. K. I.; Dietrich-Buchecker, C.
O.; Sauvage, J. P. Inorg. Chem. 1989, 28, 4070. (e) Federlin, P.; Kern,
J. M.; Rastegar, A.; Dietrich-Buchecker, C. O.; Marnot, P. A.; Sauvage,
J. P. NouV. J. Chim. 1990, 14, 9.
(5) (a) McGarvey, J. J.; Bell, S. E. J.; Bechara, J. N. Inorg. Chem. 1986,
25, 4327. (b) Bell, S. E. J.; McGarvey, J. J. Chem. Phys. Lett. 1986,
124, 336. (c) McGarvey, J. J.; Bell, S. E. J.; Gordon, K. C. Inorg.
Chem. 1988, 27, 4003.
(6) (a) Sakaki, S.; Koga, G.; Ohkubo, K. Inorg. Chem. 1986, 25, 2330.
(b) Sakaki, S.; Koga, G.; Hinikuma, S.; Hashimoto, S.; Ohkubo, K.
Inorg. Chem. 1987, 26, 1817.
(7) (a) Gamache, R. E.; Rader, R. A.; McMillin, D. R. J. Am. Chem. Soc.
1985, 107, 1141. (b) Crane, D. R.; Ford, P. C. J. Am. Chem. Soc.
1991, 113, 8510. (c) Crane, D. R.; Ford, P. C. Inorg. Chem. 1993, 32,
2391.
Experimental Section
Reagents. Copper(I) oxide (97%), 1,10-phenanthroline (abbreviated
phen) (99+%), neocuproine (2,9-dimethyl-1,10-phenanthroline, dmp)
(99+%), bathocuproine (2,9-dimethyl-4,7-diphenyl-1,10-phenanthroline,
bcp) (98+%), and anthracene (zone refined, 99%) were purchased from
Aldrich Chemical Co. Methyl viologen dichloride hydrate (MV2+
)
(97%) and benzyl viologen dichloride (BV2+) (98%) (Aldrich Chemical
Co.) were metathesized to the PF6 salts by reaction with NH4PF6 in
aqueous solution. Electrochemical grade (>99%) tetrabutylammonium
(8) Edel, A.; Marnot, P. A.; Sauvage, J. P. NouV. J. Chim. 1984, 8, 495.
(9) (a) Breddels, P. A.; Blasse, G.; Casadonte, D. J.; McMillin, D. R.
Ber. Bunsen-Ges. Phys. Chem. 1984, 88, 572. (b) Alonso-Vante, N.;
Nierengarten, J. F.; Sauvage, J. P. J. Chem. Soc., Dalton Trans. 1994,
1650.
(10) Castellano, F. N.; Ruthkosky, M.; Meyer, G. J. Inorg. Chem. 1995,
34, 3.
(11) (a) Johansen, O.; Mau, A.; Sasse, W. H. F. Chem. Phys. Lett. 1983,
94, 113. (b) Mau, A.; Johansen, O.; Sasse, W. H. F. Photochem.
Photobiol. 1985, 41, 503. (c) Olmsted, J.; McClanahan, S. F.;
Danielson, E.; Younathan, J. N.; Meyer, T. J. J. Am. Chem. Soc. 1987,
109, 3297. (d) Olmsted, J., III; Meyer, T. J. J. Phys. Chem. 1987, 91,
1649.
S0020-1669(96)00503-4 CCC: $12.00 © 1996 American Chemical Society

Copper Phenanthroline Excited States
Inorganic Chemistry, Vol. 35, No. 22, 1996 6407
hexafluorophosphate (TBAH) and tetrabutylammonium benzoate were
used as received from Fluka Chemical Co. Anhydrous diethyl ether
(Fisher) was used as received. Dichloromethane and acetonitrile
(Fisher) were distilled under argon over calcium hydride (Aldrich).
Preparations. Cu(CH3CN)4(PF6) was prepared and satisfactorily
characterized by a published procedure.12 Ru(bpy)3(PF6)2 was available
from previous studies.18
dpp. The ligand 2,9-diphenyl-1,10-phenanthroline (abbreviated dpp)
was prepared and characterized by a published procedure.13
General Preparation of Copper Phenanthroline Derivatives. A
general procedure for the synthesis of cuprous compounds of the type
Cu(NN)2(PF6) is as follows. Two equivalents (1-2 mmol) of a
phenanthroline derivative was dissolved in 10 mL of acetonitrile, and
the resultant solution was deaerated with argon. This solution was
added, with stirring, to solid Cu(CH3CN)4(PF6) under an argon
atmosphere. An orange-red color was immediately observed, and the
reaction mixture was stirred for an additional 10 min. The product
was then precipitated by adding ∼50 mL of argon-saturated diethyl
ether. The solid was filtered onto a glass frit and redissolved in a
minimum of CH2Cl2. Slow, layered addition of diethyl ether produced
air-stable crystals of the product in 88-94% yield. The solids were
stored in air until use.
amount of precipitate formed. The resultant mixture was stirred for 2
h, and an orange solid was vacuum-filtered onto a medium-porosity
glass frit. The yield was 63%. 1H NMR, δ: from dmp, 2.40 (s, 12H,
CH3), 7.97 (d, 4H, H3, H8), 8.23 (s, 4H, H5, H6), 8.77 (d, 4H, H4, H7);
from anion, 7.20 (m, 2H), 7.81 (m, 3H).
Experimental Procedures. Absorbance Measurements. Steady
state absorption measurements were performed on either an HP 8451A
diode array or a Cary 14 spectrometer. The reference was neat
dichloromethane. The apparatus used to measure excited state absor-
bance spectra was previously described.14 The second harmonic from
a Continuum Surelite Nd:YAG laser provided excitation (λex ) 532
nm; 5 mJ/pulse; fwhm 5-7 ns). Alternatively, the third harmonic was
used to pump an in-house constructed high-pressure Raman shifter filled
with H2 gas (400 psi) and the first Stokes line isolated (λex ) 416 nm;
4 mJ/pulse). A pulsed 150 W Xe lamp (Applied Photophysics) probe
beam was focused onto the entrance slit of an Applied Photophysics
f/3.4 monochromator coupled to an R446 photomultiplier tube in a
shielded housing. The instrument response function (IRF) was 20 ns.
Photoluminescence Measurements. Corrected photoluminescence
(PL) and excitation spectra were recorded on a SPEX Fluorolog, Model
1681/1680, which had been calibrated with an NBS standard lamp.
Excitation spectra were recorded by reference to a Rhodamine B
quantum counter. PL quantum yields were made with Ru(bpy)3(PF6)2
as a quantum counter in the optically dilute technique.15 The PL
quantum yield for Ru(bpy)3(PF6)2 in argon-purged water is Φr )
0.042.16 The refractive indices are 1.3387 for water and 1.424 for
dichloromethane.17
Cu(phen)2(PF6). Yield: 88%. 1H NMR ((CD3)2CO), δ: 8.08 (m,
4H, H2, H9), 8.34 (s, 4H, H5, H6), 8.89 (d, 4H, H3, H8), 9.13 (d,4H, H4,
H7). FAB-MS: m/e 423 (M+), 243 (M+ - L). Anal. Calcd for
C24H16N4CuPF6: C, 50.67; H, 2.83; N, 9.85. Found: C, 50.72; H,
2.69; N, 9.85.
Cu(dmp)2(PF6). Yield: 90%. 1H NMR (CDCl3), δ: 2.44 (s, 12H,
CH3), 7.78 (d, 4H, H3, H8), 8.02 (s, 4H, H5, H6), 8.49 (d, 4H, H4, H7).
FAB-MS: m/e 497 (M+), 271 (M+ - L). Anal. Calcd for C28H24N4-
CuPF6: C, 53.81; H, 3.87; N, 8.96. Found: C, 53.46; H, 3.66; N,
8.88.
Cu(dpp)2(PF6). Yield: 88%. 1H NMR (CDCl3), δ: 6.53 (t, 8H,
phenyl), 6.80 (t, 4H, phenyl), 7.39 (d, 8H, phenyl), 7.88 (d, 4H, H3,
H8), 8.02 (s, 4H, H5, H6), 8.51 (d, 4H, H4, H7). FAB-MS: m/e 727
(M+), 395 (M+ - L). Anal. Calcd for C48H32N4CuPF6: C, 66.02; H,
3.69; N, 6.42. Found: C, 64.68; H, 3.84; N, 6.29.
Cu(bcp)2(PF6). Yield: 94%. 1H NMR (CDCl3), δ: 2.60 (s, 12H,
CH3), 7.60 (m, 20H, phenyl), 7.76 (d, 4H, H3, H8), 8.05 (s, 4H, H5,
H6). FAB-MS: m/e 783 (M+), 423 (M+ - L). Anal. Calcd for
C52H40N4CuPF6: C, 67.21; H, 4.34; N, 6.03. Found: C, 67.46; H,
4.17; N, 6.01.
Time-resolved PL measurements were obtained on an apparatus that
was previously described.18 The IRF was 14 ns. The samples were
excited with 460 nm light.
Quenching Experiments. For quenching experiments with an-
thracene, 100 mL stock solutions of the copper phenanthroline
compounds (OD ) 0.05-0.15) in dichloromethane were prepared. In
these solutions were dissolved appropriate amounts of anthracene in 5
or 10 mL volumetric flasks. After transfer to long-neck quartz cuvettes,
these solutions were then deaerated with CH2Cl2-saturated argon for
20 min. The PL spectra, time-resolved PL decays, and the excited
state absorption spectra were recorded. Electron transfer studies were
performed in the same general manner except stock solutions of Cu-
(dpp)2(PF6) in 0.1 M TBAH/CH3CN were used. UV/vis absorbance
measurements were performed before and after quenching experiments
to ensure that no photochemistry had occurred.
Electrochemical Measurements. Cyclic voltammetry was per-
formed in a one-compartment cell with a Pt button working electrode,
a Pt gauze counter electrode, and a Ag/AgNO3 reference electrode.
The reference electrode (BAS) consisted of a silver wire in a 0.1 M
TBAH, 0.01 M AgNO3 acetonitrile solution. A BAS CV 27 potentiostat
was employed in conjunction with an HP 7035B X-Y chart recorder.
Typical solutions consisted of ∼10 mM concentrations of the copper
compounds in 0.1 M TBAH in CH2Cl2 or CH3CN.
1H NMR spectra were obtained on a Bruker AMX-300 MHz
instrument with a TMS reference. FAB mass spectra were measured
at the Johns Hopkins University mass spectrometry facility. Elemental
analyses (C, H, N) were obtained from Atlantic Microlab, Inc.,
Norcross, GA, or Desert Analytics, Tuscon, AZ.
Cu(dmp)2AnCO2. A suspension of AnCOOH (1 g) in a solution
of NaHCO3 in H2O (2.5 g/50 mL) was heated to 60 °C and stirred for
1.5 h, 20 mL of H2O was added, and the solution was evaporated to
dryness. The solid was partially dissolved in 250 mL of ethanol, the
mixture was filtered, and the filtrate was evaporated to dryness. The
product, formulated as NaAnCO2, was recrystallized from ethanol. In
a separate beaker, Cu(dmp)2(PF6) (100 mmol) was dissolved in a
minimal amount of acetone with 5 equiv of tetrabutylammonium
chloride and stirring. An additional 10 mL of acetone was added, and
the solution was placed in a freezer for 30 min. An orange solid was
vacuum-filtered onto a medium-porosity glass frit. The complex,
formulated as Cu(dmp)2Cl, was recrystallized from a minimal amount
of warm water. The Cu(dmp)2Cl was then dissolved in a minimal
amount of water with slight heating. Separately, 5.0 equiv of NaAnCO2
was dissolved in a minimal amount of water. The two solutions were
mixed with stirring, and an orange precipitate immediately formed. The
resultant mixture was stirred for 2 h, and the product Cu(dmp)2AnCO2
was vacuum-filtered onto a medium-porosity glass frit. The yield, based
on copper, was 80%. 1H NMR, δ: from dmp, 2.39 (s, 12H, CH3),
7.97 (d, 4H, H3, H8), 8.23 (s, 4H, H5, H6), 8.77 (d, 4H, H4, H7); from
anion, 7.44 (m, 4H, H2, H3, H6, H7), 7.89 (m, 4H, H1, H4, H5, H8), 8.10
(s, 1H, H9).
Results
The cuprous phenanthroline hexafluorophosphate salts were
prepared in high yields by a simple synthetic scheme. Despite
the fact that these compounds have been known for some
time,3a,4b the known lability of cuprous compounds led us to
characterize our products by proton NMR, FAB-MS, and
Cu(dmp)2(C6H5CO2). Cu(dmp)2Cl was dissolved in a minimal
amount of water with a small amount of heat. Separately, 1.5 equiv
of tetrabutylammonium benzoate was dissolved in a minimal amount
of water. The two solutions were then mixed with stirring, and a small
(14) Argazzi, R.; Bignozzi, C. A.; Heimer, T. A.; Castellano, F. N.; Meyer,
G. J. Inorg. Chem. 1995, 33, 5741.
(15) Demas, J. N.; Crosby, G. A. J. Phys. Chem. 1971, 75, 991.
(16) Caspar, J. V.; Meyer, T. J. J. Am. Chem. Soc. 1983, 105, 5583.
(17) Washburn, E. W., West, C. J., Dorsey, N. E., Ring, M. D., Eds.
International Critical Tables of Numerical Data. Physics, Chemistry,
and Technology, 1st ed.; McGraw Hill: New York, 1930; Vol. 7.
(18) Castellano, F. N.; Heimer, T. A.; Thandhasetti, M.; Meyer, G. J. Chem.
Mater. 1994, 6, 1041.
(12) Inorg Synth. 1979, 19, 90.
(13) Dietrich-Buchecker, C. O.; Marnot, P. A.; Sauvage, J. P. Tetrahedron
Lett. 1982, 23, 5291.

6408 Inorganic Chemistry, Vol. 35, No. 22, 1996
Ruthkosky et al.
Table 1. Photophysical and Redox Properties of Cuprous Phenanthroline Derivativesa
d
compound
λ
abs, nm (ꢀ)b
λ
PL, nmc
104ΦPL
τ, nse
∆Ges,f eV
E1/2ox,g V
E* ox,h V
1/2
Cu(phen)2(PF6)
Cu(dmp)2(PF6)
Cu(dpp)2(PF6)
Cu(bcp)2(PF6)
458 (6880)
454 (7950)
448 (3440)
478 (13 200)
< 20
85
250
70
0.19
0.64
0.58
0.58
740
715
765
2.3
9.7
1.5
2.04
1.99
1.98
-1.4
-1.4
-1.4
a Photophysical data were measured in neat dichloromethane. b The molar extinction coefficient at the absorption maximum is given in parentheses
Μ-1 cm-1. This data are in excellent agreement with previously reported spectra.2b c Corrected photoluminescence maximum, (4 nm. d Lower
limits for the PL quantum yields. See Results for more details. e Excited state lifetime in argon-saturated dichloromethane, (5%. f Estimated free
energy in the excited state, (50 mV. g Measured in 0.1 M TBAH/CH2Cl2, (0.01 V Vs Ag/AgNO3, which was found to be +0.35 Vs SCE using
ferrocene as an internal standard. h Excited state oxidation potential calculated using eq II in the text.
elemental analysis to ensure their purity and identity. The
compounds were characterized electrochemically and optically
as described below.
Figure 1 displays cyclic voltammograms of Cu(phen)2(PF6),
Cu(dmp)2(PF6), Cu(dpp)2(PF6), and Cu(bcp)2(PF)6 in dichlo-
romethane electrolyte solution versus the Ag/AgNO3 reference.
The observed wave is assigned to the CuII/I couple. A ferrocene
standard demonstrates that the process involves one electron.
The redox chemistry shows equivalent anodic and cathodic peak
currents, ipa/ipc ∼ 1, and a peak-to-peak separation of ∆Epp∼
100 mV. ∆Epp observed for the copper compounds is less than
or equal to that observed with the ferrocene+/0 couple in this
electrolyte. Plots of the square root of the peak current versus
scan rate are linear as expected for a diffusional process.19
Attempts to reduce the compounds led to irreversible processes
at a platinum working electrode. Electrochemical results are
summarized in Table 1.
Excitation into the broad visible absorbance bands produces
room-temperature photoluminescence (PL) for all the com-
pounds except Cu(phen)2(PF6). A lower limit for the PL
quantum yield was calculated, since the PL tails into the near-
infrared beyond the sensitivity of the photomultiplier tube.
Excitation spectra confirm that the PL is from the copper
compounds. A tangent drawn to the high-energy side of the
Figure 1. Cyclic voltammograms for the four copper diimine
compounds in 0.1 M TBAH dichloromethane solution: (a) Cu(phen)2-
(PF6); (b) Cu(bcp)2(PF6); (c) Cu(dmp)2(PF6); (d) Cu(dpp)2(PF6). The
data were collected at 200, 100, 50, and 20 mV/s. Other details can be
found in the Experimental Section and under Results.
650 to 750 nm and the laser irradiance from 40 to 60 µJ/pulse.
The lifetimes were typically within 5% from 0.01 to 0.1 mM
concentrations. The photophysical data are summarized in Table
1.
Excited state absorption difference spectra were obtained for
all compounds except Cu(phen)2(PF6), which displayed no
transient absorbance signals after 20 ns. The excited state
absorption difference spectra, obtained 20, 100, and 200 (or 500)
ns after a 532 nm laser pulse, show isosbestic points in
dichloromethane, Figure 2. Positive absorption features ob-
served at ∼340 and ∼580 nm are assigned to the reduced
ligand.21 A bleach of the visible charge transfer bands is also
observed. The absorption features decay by a first-order process
with kinetics which are the same as those obtained from time-
resolved PL measurements. The kinetics are also in good
agreement with previously reported lifetimes under similar
conditions.2b
corrected PL spectra was used to crudely estimate the free
20a
energy in the excited state, ∆Ges.
Ignoring work terms, ∆Ges
can be used to estimate the excited state oxidation potential,
ox
E* by eq II.20 Within experimental error, the excited state
1/2
E1*/2ox ) E1/2ox - ∆Ges
(II)
The addition of anthracene to a dichloromethane solution of
an emissive copper compound leads to quenching of the PL
intensity and lifetime. The concentration-dependent quenching
is well fit to the Stern-Volmer model, eq III.22 Here PLI0 is
oxidation potentials are the same for the three emissive copper
compounds, -1.4 V vs Ag/AgNO3. Time-resolved PL decays
are first order, independent of the observation wavelength from
(21) (a) Nakumura, T.; Soma, M.; Onishi, T.; Tamura, K. Z. Phys. Chem.
(Munich) 1974, 89, 122. (b) Kato, T.; Shida, T. J. Am. Chem. Soc.
1979, 101, 6869. (c) Turro, C.; Chung, Y. C.; Leventis, N.;
Kuchenmeister, M. E.; Wagner, P. J.; Leroi, G. E. Inorg. Chem. 1996,
35, 5104.
(22) Lakowicz, J. R. Principles of Fluorescence Spectroscopy; Plenum
Press: New York, 1983.
(19) Bard A. J.; Faulkner, L. R. Electrochemical Methods: Fundamentals
and Applications; Wiley: New York, 1980.
(20) (a) Arnold, D. R.; Baird, N. C.; Bolton, J. R.; Brand, J. C. D.; Jacobs,
P. W. M.; DeMayo, P. Ware, W. R. Photochemistry. An Introduction;
Academic Press: New York and London, 1974; p 13. (b) Rehm, D.;
Weller, A. Isr. J. Chem. 1970, 8, 259.

Copper Phenanthroline Excited States
Inorganic Chemistry, Vol. 35, No. 22, 1996 6409
Table 2. Copper Phenanthroline Energy Transfer to Anthracene
Φent KSV,b M-1
kQ,b M-1 s-1
kent,c M-1 s-1
a
compound
Cu(dmp)2(PF6) 0.25
Cu(dpp)2(PF6) 0.15
Cu(bcp)2(PF6) 0.10
17
7
d
(1.9 ( 0.3) × 108 (2.2 ( 0.2) × 108
(2.6 ( 0.3) × 107 (2.5 ( 0.2) × 107
d
(1.8 ( 0.2) × 107
a Quantum yield for energy transfer, relative to Ru(bpy)22+* as a
donor. b Obtained from Stern-Volmer analysis in argon-saturated
CH2Cl2. c Measured by the appearance of the triplet-to-triplet An
absorbance following selective excitation of the copper compounds in
argon saturated CH2Cl2. The error represents one standard deviation
from at least three separate trials. d The quenching was inefficient and
the Stern-Volmer model could not be quantitatively tested. If the
quenching data do fit the model, Ksv < 5 M-1
.
Figure 3. Time-resolved absorption kinetics recorded at 410 nm after
selective excitation of Cu(dmp)2(PF6) in dichloromethane solution with
0.1 M anthracene. The inset shows the excited state absorption
3
difference spectra, assigned to the triplet-to-triplet absorption of An,
recorded 2 µs after excitation with a 532 nm laser pulse.
assignment is made on the basis of previous literature reports23
and the appearance of a similar absorbance feature following
direct 355 nm excitation of anthracene.
Figure 2. Excited state absorption difference spectra obtained in neat
dichloromethane after excitation with 532 nm light: (a) for Cu(bcp)2-
(PF6), recorded at 20, 100, and 200 ns; (b) for Cu(dmp)2(PF6), recorded
20, 100, and 200 ns; (c) for Cu(dpp)2(PF6), recorded 20, 100, and 500
ns.
The relative efficiency of energy transfer was compared to
that obtained with Ru(bpy)32+* as the donor. In these experi-
ments, three argon-saturated 0.022 M anthracene/dichloro-
methane solutions were prepared and a donor was added to each
solution such that the absorbance at 532 nm was the same, A532
) 0.070 ( 0.005. The relative yield of triplet anthracene was
then measured 2 µs after 532 nm laser excitation for each
solution. A 2 µs delay was chosen to ensure that the absorbance
PLI0
PLI
τ
)
0 ) 1 + KSV[An]
(III)
τ
3
signal was solely due to An; however, delays from 100 ns to
the integrated PL intensity in the absence of anthracene, τ0 is
the excited state lifetime in the absence of anthracene, PLI and
τ are the corresponding quantities in the presence of anthracene,
and KSV is the Stern-Volmer quenching constant. Plots of PLI0/
PLI and τ0/τ are linear and coincident for Cu(dpp)2+* and
Cu(dmp)2+* quenching. Typical maximum PLI0/PLI and τ0/τ
values were between 2 and 3. Anthracene concentrations from
1 × 10-5 to 7 × 10-2 M were utilized. Cu(bcp)2+* was also
quenched by anthracene; however, the short lifetime, inefficient
energy transfer, and the low solubility of anthracene in dichlo-
romethane frustrated our attempts to obtain a reliable KSV value.
From these attempts, KSV is clearly <5 M-1. However, a much
more accurate value for the rate of energy transfer from Cu-
(bcp)2+* to anthracene was obtained by excited state absorption
spectroscopy, Table 2.
10 µs result in the same relative yields. The yields are
summarized in Table 2. The most efficient donor is Cu-
(dmp)2+*, which produces one-fourth of the amount of 3An that
Ru(bpy)32+* does under these conditions. Energy transfer from
Ru(bpy)32+* to anthracene has been shown to occur with a
quantum yield close to unity under a variety of conditions.11
The triplet-to-triplet absorbance growth reflects the kinetics
of energy transfer from the copper excited states. In our
3
previous studies, the formation of An was monitored at 430
nm, where the kinetics were complicated somewhat by contribu-
tions from unquenched copper excited states. Inefficient energy
transfer and poor solubility of anthracene precluded complete
quenching, and this contribution was subtracted. Here, we
3
monitored An at the isosbestic points shown in Figure 2 and
could cleanly observe the energy transfer process. We note that
the kinetics measured at the isosbestic wavelength are in
excellent agreement with those reported previously10 and plots
of the observed rate versus anthracene concentration are linear.
The appearance of the triplet-to-triplet absorption of triplet
3
anthracene, An, centered at 430 nm following selective laser
excitation of the copper phenanthroline compounds directly
demonstrates energy transfer, Figure 3. All the copper phenan-
throline compounds except Cu(phen)2(PF6) transfer energy to
-
Cu(dmp)2(AnCO2), where AnCO2 is the 9-anthracenecar-
boxylate anion, and Cu(dmp)2(C6H5CO2), which serves as a
3
anthracene. No An absorption is observed in the absence of
(23) Birks, J. B. Photophysics of Aromatic Molecules; John Wiley & Sons
the copper compounds under these conditions. The 3An
Ltd.: London, 1970; p 258.

6410 Inorganic Chemistry, Vol. 35, No. 22, 1996
Ruthkosky et al.
+
Table 3. Electron Transfer of Cu(dpp)2 Viologens in 0.1 M
TBAH/CH3CN
10-9kf,b
10-10kb,d
c
acceptora -∆Gf, b V M-1 s-1
ΦCE
-∆Gb,d M-1 s-1
MQ+
MV2+
BV2+
0.20
0.55
0.63
4.3 ( 0.5
6.6 ( 0.5 0.95 ( 0.05
6.7 ( 0.5 0.57 ( 0.05
2.0
1.4
1.3
2.0 ( 0.1
2.1 ( 0.1
a Electron acceptors where MQ+ is monoquat, MV2+ is methyl
viologen, BV2+ is benzyl viologen. b Driving force and electron for
electron transfer from Cu(dpp)2+* to the viologens calculated as
described in the Results. c Cage escape yields calculated with eq IV.
The error represents one standard deviation from three separate
measurements. d Driving force and rates for back electron transfer from
2+
the reduced viologen acceptor to Cu(dpp)2
.
the estimated -1.4 V excited state oxidation potential of
Cu(dpp)2+* and the reduction potential of the acceptors
measured in the same electrolyte by cyclic voltammetry. The
reduction potentials were measured to be -0.85 V for MV2+/+•
,
-0.77 V for BV2+/+•, and -1.38 V for MQ+/0 under these
conditions. The back electron transfer processes follow second-
order equal concentration kinetics (inset, Figure 4b). The kinetic
data are summarized in Table 3. Cage escape efficiencies with
[Cu(dpp)22+, MV+•] were determined in samples where the
fraction of PL quenched was typically >95%. Actinometry was
performed before and after each experiment by measuring the
2+
MLCT-state transient absorbance of an aqueous Ru(bpy)3
solution, assuming a quantum yield of unity for formation of
the MLCT excited state and ∆ꢀ360 ) 2.2 × 104 M-1 cm-1. The
changes in extinction coefficient (∆ꢀ) for MV+• at λ ) 396
Figure 4. (a) Excited state absorption difference spectrum of a Cu-
(dpp)2(PF6) acetonitrile solution with 0.1 M TBAH and 10 mM MV2+
recorded 5 µs after a 532 nm laser pulse. (b) Time-resolved absorbance
measured at 396 nm. The inset shows the fit of these same data to a
second-order equal-concentration kinetic model.
and 606 nm were taken to be 41 800 and 13 900 M-1 cm-1
,
respectively.24 Cage escape yields (ΦCE) were then calculated
from the absorbances of the sample and actinometer solution
at 416 or 532 nm, eq IV, in conjunction with percent PL
model, were prepared by metathesis reactions. Integration of
the proton NMR confirms that the cation and anion are present
in a 1:1 molar ratio. The lifetimes and PL intensities are highly
concentration dependent. At low concentrations, the lifetimes
reach a limiting upper value which decreases rapidly with
increased concentration into our instrument response function,
14 ns. At low concentrations, the lifetimes of the AnCO2- and
C6H5CO2- salts are comparable to the lifetime of the PF6- salt,
τ ) 80 ( 10 ns. Using the low-concentration lifetime as τ0,
plots of τ0/τ are linear and give Stern-Volmer constants of 4000
M-1 for Cu(dmp)2(C6H5CO2) and 4400 M-1 for Cu(dmp)2-
(AnCO2). The excited state absorption spectrum is essentially
identical to that of Cu(dmp)2(PF6), with kinetics which agree
well with PL data. No absorption feature due to 3AnCO2- was
observed under any conditions.
Electron transfer from copper diimine excited states to methyl
viologen (MV2+), benzyl viologen (BV2+), and monoquat
(MQ+) were explored in CH3CN/0.1 M TBAH electrolyte.
Acetonitrile was chosen due to the poor solubility of the
viologen salts in dichloromethane. The only compound which
is emissive in CH3CN/0.1 M TBAH electrolyte is Cu(dpp)2-
(PF6), which has a lifetime of 110 ns and a corrected PL
maximum at 750 nm. In all cases, electron transfer quenching
(1 - 10-A
(1 - 10-A
)
actinometer
532 or 416
(∆Aλ/∆ꢀλ)A
-
ΦCE
)
×
+
Cu(dpp)2
532 or 416
(∆A360/∆ꢀ360 actinometer
)
)
1
(IV)
fraction of PL quenched
quenched determined from extrapolation of Stern-Volmer plots.
For five different samples, ΦCE was 0.95 ( 0.05. In contrast,
ΦCE for [Ru(bpy)33+, MV+•] was measured to be 0.25 ( 0.05
under the same conditions, in agreement with literature values.25
A series of experiments were conducted to verify the higher
cage escape efficiency for the copper compound. TBAH (0.1
2+
M) solutions with 10 mM MV2+ and either Ru(bpy)3 or
Cu(dpp)2+ were prepared so that the absorbance at 532 (or 416)
nm was the same, typically 0.05 (or 0.3). The relative amount
of MV+• formed after the laser pulse was then monitored by
excited state absorption spectroscopy. The yield was typically
3 times greater for Cu(dpp)22+. The cage escape yield for [Cu-
(dpp)22+, BV+•] was measured analogously with the known
extinction coefficient of BV+•.26
Discussion
was observed by the appearance of the reduced viologen after
Cuprous phenanthroline compounds are photostable and
display broad charge transfer bands in the visible region,
potentially useful for solar harvesting. In some cases, these
excited states sensitize the reduction of viologens and/or energy
transfer to anthracene. Below, we discuss the excited state and
+
selective excitation of Cu(dpp)2
.
Electron transfer to MV2+ after selective excitation of
+
Cu(dpp)2 was explored in most detail, Figure 4.24 Plots of
the observed electron transfer rates monitored at 396 nm versus
the viologen concentrations are linear. The driving forces for
the electron transfer processes were calculated on the basis of
(25) Hoffman, M. Z. J. Phys. Chem. 1988, 92, 3458 and references therein.
(26) Yonemoto, E. H.; Riley, R. L.; Kim, Y. I.; Atherton, S. J.; Schmehl,
R. H.; Mallouk, T. E. J. Am. Chem. Soc. 1992, 114, 8081.
(24) Watanabe, T.; Honda, K. J. Phys. Chem. 1982, 86, 2617.

Copper Phenanthroline Excited States
Inorganic Chemistry, Vol. 35, No. 22, 1996 6411
Chart 1
redox properties of the compounds, followed by the interesting
energy and electron transfer processes.
While the demonstration of energy transfer is clear, the
efficiency is low in all cases. Relative to that of Ru(bpy)32+*,
which is known to transfer energy to anthracene derivatives with
a quantum yield near unity under a variety of conditions,11 the
yield from the copper excited states is at best 0.25. Cu(dmp)2-
(AnCO2) was prepared with the hope that ion-pairing and,
possibly, exciplex formation would mediate energy transfer and
thereby increase the efficiency.31,32 While our results reveal
some indirect evidence of excited state complex (exciplex)
formation, there is no spectroscopic evidence of energy transfer.
Instead, the concentration-dependent quenching can be satis-
factorily rationalized on the basis of an ion-pair exciplex model
proposed by McMillin in which counterion coordination to the
Cu center promotes nonradiative decay. A fundamental predic-
tion of this model is that the extent of quenching should mirror
the nucleophilicity of the anion. A comparison of published
Excited State and Redox Properties. All four copper
compounds are reversibly oxidized in dichloromethane elec-
trolyte. A strikingly more negative oxidation potential for
+
Cu(phen)2 compared to the other copper compounds is
observed. Williams and James first reported that the oxidation
+
potential of Cu(phen)2 is 420 mV more negative than that of
+
Cu(dmp)2 in water.27 The results here show a similar effect
in a noncoordinating solvent. The observation cannot be
rationalized by electronic considerations but can be based on
+
steric interactions. The known crystal structures of Cu(dmp)2
2+
and Cu(dmp)2 show that the cupric state places the methyl
groups in the 2- and 9-positions of the ligand into an unfavorable
geometry, Chart 1.28 Therefore, the presence of substituents in
these positions favors the tetrahedral cuprous state. The slightly
more positive oxidation potential for Cu(bcp)2+ can be rational-
ized on the basis of addition of the phenyl substituents.29 We
note that, despite the large inner-sphere reorganization energy
which accompanies this redox chemistry, the voltammetry
observed is nearly ideal.
data to these indicates that the nucleophilicity of anions
-
decreases in the order PF6- > AnCO2- g C6H5CO2- > NO3
.
The ion-pair exciplex quenching process deactivates the excited
state rather than providing a conduit for energy transfer.
Electron Transfer. Electron transfer from Cu(dpp)2+* to
three viologens, kf, is clearly observed in the reduced viologen
absorption spectra. The increased rate for MV2+ and BV2+
when compared to MQ+ is consistent with electron transfer in
the normal region, Table 3. The yield of MV+• observed after
the laser pulse is 3-4 times higher than that observed when
Ru(bpy)32+* is the donor. The increased yield can be traced to
a cage escape yield near unity within the [Cu(dpp)22+, MV+•]
charge-separated pair. Previous studies have shown that spin
changes are a key factor in determining cage escape efficiencies,
and they do explain the results here. Triplet ion pairs yield
higher cage escape efficiencies than do singlet ion pairs.11d,33
For the [Ru(bpy)33+, MV+•] pair, spin-orbit coupling increases
the singlet character and the cage escape yield drops in
comparison to the case of an organic triplet donor, such as
3An.11d For the copper donors, spin-orbit coupling likely plays
a lesser role and the excited state is closer to a pure triplet state,
which results in a triplet ion pair and a high cage escape yield.
An alternative explanation is that a structural barrier exists for
recombination within the [Cu(dpp)22+, MV+•] solvent cage since
different geometries are expected for the forward and reverse
electron transfer processes. In either case, high cage escape
yields may be a general phenomena for chromophoric copper
diimine donors. In this regard, we note that the cage escape
yield for [Cu(dpp)22+, BV+•] is also very high. Recombination
of the charge-separated pairs to ground state products, kb, occurs
within experimental error of the diffusion limit. The diffusion
The broad visible absorption bands were previously attributed
to metal-to-ligand charge transfer (MLCT) excited states.2
+
Excited state absorption spectroscopy of Cu(dmp)2+, Cu(dpp)2
,
+
and Cu(bcp)2 supports this assignment. In the one-electron
transfer localized extreme, the excited state can be viewed as a
CuII metal center coordinated to a reduced ligand. Difference
spectra show isosbestic points consistent with the formation of
a reduced ligand and CuII metal.21 Further, the dramatic red
shift in the PL spectra can be taken as evidence for emission
from a geometrically relaxed Cu(II) excited state. The lack of
+
a resolvable excited state absorption spectrum for Cu(phen)2
is indicative of rapid nonradiative decay, knr > 5 × 107 s-1
,
which is expected on the basis of the energy gap law.30
Energy Transfer. Dynamic triplet-triplet energy transfer
from the long-lived copper MLCT excited states to anthracene
is observed. The energy transfer rates measured by Stern-
Volmer analysis of the PL quenching and the appearance of
the triplet-to-triplet 3An absorption are in excellent agreement.
+
The order-of-magnitude-faster energy transfer from Cu(dmp)2
*
can be rationalized on the basis of ∼50 mV larger driving force,
Table 1. The lack of energy transfer from Cu(phen)2+* is likely
kinetic in origin, since these excited states do not live long
enough to undergo diffusional quenching.
(27) James, B. R.; Williams, J. P. J. Chem. Soc. 1961, 2007.
(28) The compounds were generated with Chem-Draw 3D from the reported
crystal structure coordinates. For clarity, the nitrate adduct in the cupric
compound was eliminated from Chart 1. (a) Dessy, G.; Fares, V. Cryst.
Struct. Commun. 1979, 8, 507. (b) Van Meersche, M.; Germain, G.;
Declercq, J. P.; Wilputte-Steinert, L. Cryst. Struct. Commun. 1981,
10, 47.
(31) Everly, R. M.; McMillin, D. R. Photochem. Photobiol. 1989, 50, 711.
(32) Fujita, I.; Kobayashi, H. J. Chem. Phys. 1973, 59, 2902.
(33) Harriman, A.; Porter, G.; Wilowska, A. J. Chem. Soc., Faraday Trans.
2 1983, 79, 807.
(29) Lever, A. B. P. Inorg. Chem. 1990, 29, 1271.
(30) Meyer, T. J. Prog. Inorg. Chem. 1983, 30, 389.

6412 Inorganic Chemistry, Vol. 35, No. 22, 1996
Ruthkosky et al.
limit in acetonitrile at 25 °C based on the Smoluchowski-
Conclusions
Stokes-Einstein equation is 1.9 × 1010 M-1 s-1 34
.
Dynamic electron and energy transfers have been clearly
demonstrated for cuprous phenanthroline derivatives. Cu(phen)2-
(PF6) has an excited state lifetime which is too short to undergo
diffusional quenching processes, τ < 20 ns. The rate and
efficiency of energy transfer to anthracene or electron transfer
to viologens are comparable with those for Ru(bpy)32+. The
ability to sensitize these processes to visible light with simple
inexpensive copper diimines is in itself noteworthy. An apparent
advantage of copper phenanthroline excited states over those
of Ru(bpy)32+* is the high cage escape yields observed after
oxidative quenching by viologens.
The conclusion of a cage escape yield near unity stands
somewhat in conflict with the results of Edel et al.8 These
workers found a greater than 60-fold increase in MV+• produc-
+
tion when anthracenecarboxylate was added to a Cu(dpp)2
/
ethanol solution which contained triethanolamine, TEOA, as a
sacrificial electron donor. This observation was later interpreted
2+
as due to a cage escape yield of less than 0.02 for [Cu(dpp)2
,
MV+•].11d An alternative explanation, in accord with the results
presented here, is that the increased yield of MV+• is due to
more efficient oxidation of TEOA by AnCO2 than by Cu-
•
(dpp)22+. We note also that TEOA oxidation products are
known to undergo subsequent dark reactions, which can
artificially increase the quantum yield of MV+• to 2.35 In either
case, the increased yield of MV+• upon addition of anthra-
cenecarboxylate anion is not due to a low cage escape yield for
the [Cu(dpp)22+, MV+•] pair.
Acknowledgment. We thank the National Renewable En-
ergy Laboratory (Grant NREL XAD-3-12113-04) and the
National Science Foundation (Grants CHE-9322559 and CHE-
9402935) for support of this research.
Supporting Information Available: Visible absorption and pho-
toluminescence spectra (2 pages). Ordering information is given on
any current masthead page.
(34) Murov, S. L.; Carmichael, I.; Hug, G. L. Handbook of Photochemistry,
2nd ed.; Marcel Dekker, Inc.: New York, 1993; pp 207-208.
(35) Sutin, N.; Creutz, C. Pure Appl. Chem. 1980, 52, 2717.
IC960503Z